(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "The Biological bulletin"

A? 



THE 



BIOLOGICAL BULLETIN 



PUBLISHED BY 

THE MARINE BIOLOGICAL LABORATORY 

Editorial Board 



JOHN M. ANDERSON, Cornell University J. LOGAN IRVIN, University of North Carolina 

DAVID W. BISHOP, Carnegie Institution of L . H. KLEINHOLZ, Reed College 

Washington 

JAMES CASE, University of California, J OHN H " LOCHHEAD, University of Vermont 

Santa Barbara ROBERTS RUGH, Columbia University 

JOHN W. GOWEN, Iowa State College BERTA SCHARRER, Albert Einstein College of 
SALLY HUGHES-SCHRADER, Duke University Medicine 

LIBBIE H. HYMAN, American Museum of WM. RANDOLPH TAYLOR, University of 

Natural History Michigan 

DONALD P. COSTELLO, University of North Carolina 
Managing Editor 



VOLUMI: 124 
FEBRUARY TO JUNE, 1963 




Printed and Issued by 

LANCASTER PRESS, Inc. 

PRINCE 8 LEMON STS. 

LANCASTER, PA. 



11 



THE BIOLOGICAL BULLETIN is issued six times a year at the 
Lancaster Press, Inc., Prince and Lemon Streets, Lancaster, Penn- 
sylvania. 

Subscriptions and similar matter should be addressed to The 
Biological Bulletin, Marine Biological Laboratory, Woods Hole, 
Massachusetts. Agent for Great Britain: Wheldon and Wesley, 
Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, 
W. C. 2. Single numbers $2.50. Subscription per volume (three 
issues), $6.00. 

Communications relative to manuscripts should be sent to the 
Managing Editor, Marine Biological Laboratory, Woods Hole, 
Massachusetts, between June 1 and September 1, and to Dr. 
Donald P. Costello, P.O. Box 429, Chapel Hill, North Carolina, 
during the remainder of the year. 



Second-class postage paid at Lancaster, Pa. 



LANCASTER PRESS, INC., LANCASTER, PA. 



CONTENTS 




No. 1. FEBRUARY, 1963 

PAGE 

BEERS, C. DALE 

Relation of feeding in the sea urchin Strongylocentrotus droebachiensis 

to division in some of its endocommensal ciliates 1 

BELAMARICH, FRANK A. 

Biologically active peptides from the pericardia! organs of the crab 
Cancer borealis 9 

BOWEN, SARANE THOMPSON 

The genetics of Artemia salina. II. White eye, a sex-linked mutation 17 

FlNGERMAN, MlLTON, AND CHITARU OCURO 

Chromatophore control and neurosecretion in the mud shrimp, Upogebia 
affinis 24 

GOODBODY, IVAN 

The biology of Ascidia nigra (Savigny). II. The development and sur- 
vival of young ascidians 31 

GORDON, MALCOLM S. 

Chloride exchanges in rainbow trout (Salmo gairdneri) adapted to differ- 
ent salinities 45 

HAGSTROM, BERNDT E. 

The effect of lithium and o-iodosobenzoic acid on the early development 

of the sea urchin egg 55 

LACHANCE, LEO E., AND SARAH B. BRUNS 

Oogenesis and radiosensitivity in Cochliomyia hominivorax (Diptera: 
Calliphoridae) 65 

McLsoo, D. G. R., AND STANLEY D. BECK 

Photoperiodic termination of diapause in an insect 84 

PAPI, F., AND L. PARDI 

On the lunar orientation of sandhoppers (Amphipoda Talitridae) 97 

No. 2. April, 1963 

ABE, SHIGEMI 

The effect of p-chloromercuribenzoate on amoeboid movement, rlagellar 
movement and gliding movement 107 

AUSTIN, C. R. 

Fertilization in Pectinaria ( = Cistenides) gouldii 115 

BARTH, LESTER G., AND LUCENA J. BARTH 

The relation between intensity of inductor and type of cellular differen- 
tiation of Rana pipiens presumptive epidermis 125 

iii 






IV CONTENTS 

BOYD, CARL M. AND MARTIN W. JOHNSON 

Variations in the larval stages of a decapod crustacean, Pleuroncodes 
planipes Stimpson (Galatheidae) 141 

CHEVERIE, J. CHARLES, AND W. GARDNER LYNN 

High temperature tolerance and thyroid activity in the teleost fish, Tani- 
chthys albonubes 153 

DEAN, DAVID, AND PHYLLIS A. HATFIELD 

Pelagic larvae of Nerinides agilis (Verrill) 163 

DIMOND, SISTER MARIE THERESE 

The relation between whole-body I 131 uptake to thyroid activity in the 
developing dogfish, Scyliorhinus canicula (L.) 170 

GIESE, A. C., AND A. FARMANFARMAIAN 

Resistance of the purple sea urchin to osmotic stress 182 

ROBERTS, HENRY S. 

Factors of equatorial contraction and polar membrane expansion in endo- 
sperm cytokinesis 193 

ROEDER, KENNETH D. 

Echoes of ultrasonic pulses from flying moths 200 

ROSENBAUM, ROBERT M., AND BRUCE DITZION 

Enzymic histochemistry of granular components in digestive gland cells 

of the Roman snail, Helix pomatia 211 

WERNTZ, HENRY O. 

Osmotic regulation in marine and fresh-water gammarids (Amphipoda) 225 

No. 3. JUNE, 1963 

CLONEY, RICHARD A. 

The significance of the caudal epidermis in ascidian metamorphosis. . . . 241 

COSTLOW, JOHN D., JR. 

Molting and cyclic activity in chromatophorotropins of the central nervous 
system of the barnacle, Balanus eburneus 254 

EDGAR, ARLAN L. 

Proprioception in the legs of phalangids 262 

IWASAKI, HIDEO, AND CHIKAYOSHI MATSUDAIRA 

Observation on the ecology and reproduction of free-living Conchocelis 

of Porphyra tenera 268 

LEWIS, JOHN B. 

Enviromental and tissue temperatures of some tropical intertidal marine 
animals 277 

OSBORNE, PAUL J., AND A. T. MILLER, JR. 

Acid and alkaline phosphatase changes associated with feeding, starva- 
tion and regeneration in planarians 285 

PIPA, RUDOLPH L. 

Studies on the hexapod nervous system. VI. Ventral nerve cord short- 
ening; a metamorphic process in Galleria mellonella (L.) (Lepidoptera, 
Pyrallidae) 293 

RUGH, ROBERTS, AND MARLIS WOHLFROMM 

X-irradiation-induced congenital anomalies in hybrid mice 303 



CONTENTS v 

Sfll.MIDT-KoENIG, Kl.Al S 

Sun compass orientation <>1 pigeons upon equatorial and trans-equatorial 
displacement 311 

STEINBACH, H. Bi RR 

Sodium, potassium and chloride in selected hydroids 322 

STEINBERG, SONIA X. 

The regeneration of whole polyps from ectodermal fragments of scyphis- 
toma larvae of Aurelia aurita 337 

WHITFORD, WALTER G., AND VICTOR H. HUTCHISON 

Cutaneous and pulmonary gas exchange in the spotted salamander, Am- 
bystoma maculatum 344 

WILLIAMS, CARROLL M. 

The juvenile hormone. III. Its accumulation and storage in the ab- 
domens of certain male moths 355 

ZWILLING, EDGAR 

Formation of endoderm from ectoderm in Cordylophora 368 



Vol. 124, No. 1 February, 1963 

THE 

BIOLOGICAL BULLETIN 

PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY 



RELATION OF FEEDING IN THE SEA URCHIN STRONGYLOCEN- 

TROTUS DROEBACHIENSIS TO DIVISION IN SOME OF 

ITS ENDOCOMMENSAL CILIATES l 

C. DALE BEERS 

Department of Zoolooy, University of North Carolina, Chapel Hill, North Carolina, 
and the Mount Desert Island Biological Laboratory, Salisbury Cove, Maine 

Seven species of ciliates, which are generally regarded as commensals, have been 
reported from the digestive tract of Strongylocentrotus droebachiensis in the Mt. 
Desert Island region (Powers, 1933a; Beers, 1948). Four of them (Cyclidium 
stercoris, Plagiopyla minuta, Euplotes balteatus, and Trichodina sp.) are of erratic 
occurrence and are excluded from the present study, which is therefore based on 
the remaining three, namely, the holotrichs Entodiscus borealis (Hentschel), Mad- 
senia indomita (Madsen), and Biggaria gracilis (Powers). They are of almost 
invariable occurrence in any urchin whose test exceeds 10 mm. in diameter, and the 
number of individuals of any one of them in an urchin 25 mm. in diameter or larger 
may be enormous. They were present in all the urchins of the present study. 

Although the food of the three has not been studied critically, the evidence 
indicates that it is largely bacterial. In some gastrioles of E. borealis, Powers 
(1933b, p. 129) reported "rod-like bodies which resemble bacteria," although in 
others he noted objects which look like the nuclei of intestinal epithelium cells. In 
B. gracilis he reported "bacteria and bits of algae" (1933a, p. 112), but the nature 
of the food of M. indomita was unmentioned. I have observed rod-like structures 
and short filaments, which were undoubtedly bacteria, in the gastrioles of all three 
ciliates, but some unidentified material was also present. 

Various aspects of the autecology of the ciliates have been treated elsewhere 
(Beers, 1948, 1961). Findings pertinent to the present study and reported in 1948 
may be summarized as follows. Biggaria gracilis is essentially an inhabitant of the 
rectum, which, except in drastically starved urchins, always contains food remnants. 
In the usual adequately fed urchin E. borealis occurs primarily in the stomach and 
M. indomita in the intestine. The regions of the gut being ill-defined, the distribu- 
tion of the ciliates is not a rigid one. If an urchin is kept without food for a week 
or longer, the stomach and intestine gradually become empty and the distribution of 
E. borealis and M. indomita changes, in that they shift toward the rectum, where 
they mingle with B. gracilis. Dividing specimens of B. gracilis can be found in 

1 Aided by a grant from the Research Council of the University of North Carolina. 

1 
Copyright 1963, hy the Marine Biological Laboratory 



C. DALE BEERS 

practically any urchin that is collected in nature, but it is a remarkable fact that 
dividing specimens of E. borealis and M. indomita are extremely difficult to find, 
even in ciliate populations of great density. Of 182 urchins that contained immense 
numbers of both ciliates, only six had dividing specimens of E. borealis and only 
three those of M. indomita. (Powers, 1933b, likewise had difficulty in finding 
dividing forms of E. borealis and indeed found only three in all his material, which 
was evidently plentiful.) It was concluded that division occurs cyclically in the 
two ciliates : that "short periods of intense divisional activity . . . alternate with long 
periods of non-divisional life" (p. 111). Since the urchins appeared to be ade- 
quately fed and to contain ample bacteria to support the ciliates, the conclusion 
implies that the cycles are inherent in the ciliates. 

I was never entirely satisfied with the conclusion, for cycles of the kind postu- 
lated are generally absent in ciliates. When interruptions of division occur, they 
are usually associated with environmental inadequacies or with special physiologi- 
cal states, such as senescence, autogamy, conjugation, or the production of resting 
cysts. Since the feeding habits of the urchins were uncontrolled in my earlier study, 
I returned to the subject in the summer of 1962, in an effort to ascertain whether 
the division of the ciliates is in any way related to the food of the host. More pre- 
cisely, the study concerns the following questions, the first two of which are some- 
what preliminary, whereas the third is the principal one. (1) What is the condi- 
tion of the ciliates with respect to division in urchins that have been deprived of 
food for a considerable period, meaning about two weeks? Earlier observations 
(unpublished) indicated that dividing forms of E. borealis and M. indomita are 
absent in such urchins, but the condition of B. gracilis was unrecorded in my notes. 
(2) Hence, are dividing forms of B. gracilis also absent? It is logical to expect 
a cessation of division in the absence of food, and this expectation was readily con- 
firmed. (3) Then, if these starved urchins are again supplied with ample food, is 
division resumed in B. gracilis and is a cycle of division initiated in E. borealis 
and M. indomita? 

MATERIALS AND METHODS 

From time to time during the summer, specimens of S. droebachiensis were 
collected at low tide from the rocks and shallow water near the Mt. Desert Island 
Laboratory. In size they varied from 25-70 mm. in diameter (test, excluding 
spines) ; most of them measured 40-55 mm. across. Injured specimens (spines 
damaged or integument abraded) were rejected, for these may be attacked and 
eaten by healthy ones. The urchins were transferred in groups of 60-70 to aquaria 
through which sea water flowed constantly. In these, they were kept without food 
for two weeks or longer (usually 14-16 days). Voided fecal pellets, which may be 
re-ingested, were removed daily by siphoning. 

Several representative urchins (usually 5 to 10) of each group were opened 
and examined as soon as collected, and the condition of their ciliates was recorded. 
In all the examinations of the entire study the relative abundance of each ciliate 
in the stomach, intestine, and rectum was estimated as formerly and the respective 
densities of population were recorded as "light," "moderate," or "heavy" (Beers, 
1948). If dividing individuals of a particular species were present, the number of 
such individuals, as well as the total number, was counted in samples containing 



CILIATES OF STRONGYLOCENTROTUS 3 

50-150 ciliates of the species. From these totals the percentage of dividing indi- 
viduals in the samples was calculated and this figure was assumed to represent 
conditions in the urchin. The presence of an occasional dividing specimen was 
ignored. But if one specimen or a larger number was dividing in every 100 indi- 
viduals of the species, division was judged to be significant and the appropriate 
percentage was recorded. 

After two weeks of starvation, several urchins of each group were again ex- 
amined and the condition of their ciliates was recorded. Then, smaller numbers 
of starved urchins (usually 10 to 20) were removed to other aquaria and were 
supplied with pieces of healthy fronds of the kelp Laminaria sp., which is a preferred 
food of S. droebachicnsis and is indeed adequate, as Swan (1961) demonstrated, to 
maintain it in a healthy, growing condition for at least a year. Usually the urchins 
were placed directly on, or in contact with, a piece of kelp ; thus, they sensed its 
presence at once and began to feed without delay. Since an excess of kelp was 
maintained in the aquaria, ample food was always available to the urchins. These 
laminaria-fed urchins were then opened and examined at various hourly intervals 
following the addition of food. By varying the time of day at which laminaria was 
initially supplied, examinations could be conveniently planned for any sequence of 
hourly intervals within the period of experimental feeding, which was arbitrarily 
restricted to 5 days. By using sufficient numbers of urchins, nearly all the hourly 
intervals of the 5-day period were finally represented in the results. It is true 
that the number of urchins examined for any particular hour was relatively small 
(usually two to five), since the examination of an urchin requires 15-45 minutes, 
depending on the condition of its ciliates. But the total number of such urchins 
was considerable and amounted finally to 294. 

RESULTS 

1. Condition of the ciliates in urchins collected in nature 

Thirty-two urchins taken from the various collections and examined immedi- 
ately showed the usual ciliate distribution and population densities (some "light," 
but many "moderate" to "heavy"), in agreement with earlier observations (Beers, 
1948). Whereas B. gracilis was dividing in all the urchins, only one contained 
dividing specimens of E. borealis, and another those of M. indomita, though divi- 
sion in each case was sparse. The digestive tracts of the urchins appeared to con- 
tain ample food, chiefly masses of filamentous green algae and fragments of 
laminaria. These observations indicated that the urchins which were subjected 
to starvation were healthy, adequately fed, and typical with respect to ciliates; 
that B. gracilis was dividing in all of them; but that only about 3% of them con- 
tained dividing specimens of E. borealis, and a like percentage those of M. indomita. 

2. Condition of the ciliates in starved urchins 

Thirty-five urchins from various experimental groups were examined after 
14-21 days of starvation. In all of them the stomach and the greater part of the 
intestine were empty (lacked solid food) ; only the terminal quarter of the in- 
testine and the rectum contained some undigested or indigestible material, but the 



C. DALE BEERS 

amount was small. Entodiscus bor calls was present in the intestine and rectum, as 
well as the stomach, and M. indomita was present in the rectum, as well as the in- 
testine. Evidently these ciliates, as digestion proceeded, had dispersed aborally 
from their preferred sites. In accordance with expectations, dividing specimens of 
the two were absent and population densities were reduced to "moderate" or "light." 
(All urchin ciliates escape regularly in limited numbers among the fecal pellets; in 
the absence of division, reduced numbers are the rule.) As usual, B. gracilis was 
restricted to the rectum, but dividing specimens were absent in all the urchins 
except one, in which only two such specimens could be found. Microscopic ex- 
amination of representatives of the three species showed that the cytoplasm was 
very transparent and contained relatively few gastrioles. Evidently a two-week 
period of starvation of the host was adequate to reduce the division rate of all the 
ciliates practically to zero, and it was concluded that the starved urchins contained 
to all practical purposes no dividing ciliates. 

3. Feeding habits of starved urchins 

The starved urchins fed readily on laminaria. The stomach was well filled after 
6-8 hours of feeding; the stomach and intestine after 15-18 hours (counting from 
the beginning) ; and the stomach, intestine, and rectum after about 20 hours. Feed- 
ing continued (by night as well as by day) for 60-72 hours, when many of the 
urchins evidently became surfeited and moved away from the food. On the fourth 
and fifth days of the experiment, some urchins were always feeding, while others 
rested near-by on the sides of the aquarium. Thus, feeding occurred irregularly 
on these days. Some of the urchins were kept a total of 6 to 14 days with the 
laminaria well beyond the end of the formal experimental period. These urchins 
also fed irregularly, probably as urchins feed under natural conditions. The in- 
testine and rectum were always well filled, but the amount of food in the stomach 
was variable. 

4. Condition of tJie ciliates in urchins fed laminaria after two weeks without food 

The results of the feeding experiments are summarized in Table I. It is under- 
stood that this table is a composite of all experiments, rather than a continuous 
record of a single vast experiment. The table concerns 294 urchins (total of 
Column 2), and it is obviously impossible for one investigator to start with this 
number of starved urchins and to continue the examinations uninterruptedly for 120 
hours. In the table the successive hours are grouped by 8-hour periods, and for 
reference purposes the periods are numbered (Column 1). It is important to note 
that all the urchins of any period are represented in each of the three columns 
under "Number (and percentage) of urchins. ..." A consideration of some of the 
periods will clarify the method of presentation. When two paragraphs appear 
under a period in the following account, the first deals with results presented in the 
table, and the second contains explanatory or supplementary comments. 

Period 1. There was examined a total of 20 urchins (Column 2) that had fed 
from 1 hour to 8 hours on laminaria (Column 3). Since no dividing ciliates could 
be found in any of these urchins, it follows that the remaining columns of the 
period read zero. 



CILIATES OF STRONGYLOCENTROTUS 

Period 2. In this period 30 urchins that had fed from 9 to 16 hours were ex- 
amined. In 18 of the 30, or 60% of them, E. borcalis was dividing (Column 4), 
but dividing specimens were absent in the remaining ciliates. 

The number of dividing specimens of E. borcalis in the 18 urchins usually 
amounted to 4-8% of the total, but in some urchins the number attained 15-20%. 
For example, on July 11 ten samples were examined from the stomach of a 51-mm. 
urchin that had fed for 10 hours. They contained a total of 522 individuals 
of E. borealis, of which 88 (nearly 17%) were judged by either of two criteria to 
be dividing : an elongated condition of the macronucleus or the presence of a 
transverse cytosomal constriction. Such specimens are easily recognized in living 
material, even with magnification as low as 10-20 X, for E. borealis is a large ciliate, 
measuring on the average 143 //, X 87 p. (Powers, 1933a). Many additional indi- 
viduals were evidently preparing to divide, in view of their large size (length, 
160-180 /A), and many had already divided, judged by the great variation in their 
size (length, 120-170 ju). (The individuals of stable, non-dividing populations 
of E. borcalis are remarkably uniform in size.) Microscopic examination of some 
of the specimens showed that they had lost much of their transparency and con- 
tained great numbers of gastrioles. It is a fact that this single urchin contained 
more dividing individuals of E. borealis than I had seen in all the hundreds of 
urchins examined in three earlier summers. In a 42-mm. urchin examined on 
August 16 after 15 hours of feeding, 20% of the individuals of E. borealis were 
dividing. 

Period 3. Nearly half the urchins (48%) contained dividing specimens 
of E. borealis, but M. indomita remained non-divisional. However, the division 
of B. gracilis was resumed in 40% of the urchins. 



TABLE I 

Incidence of division of three species of ciliates in urchins (Strongylocentrotus droebachiensis} 
which were starved for two weeks and then fed generously on Laminaria during a 

5-day experimental period 



Successive 
8-hour 
periods 


No. of urchins 
examined per 
8-hour period 


Time in hours 
after beginning 
of feeding 


Number (and percentage) of urchins in which ciliate 
indicated was dividing 


Enlodiscus 

borealis 


Madsenia 
indomita 


Biggaria 
gracilis 


1 


20 


1-8 











2 


30 


9-16 


18 (60) 








3 


25 


17-24 


12 (48) 





10 (40) 


4 


21 


25-32 


10 (48) 





17 (81) 


5 


19 


33-40 


9 (47) 





19 (100) 


6 


22 


41-48 


(41 ) 





22 (100) 


1 


17 


49-56 


7 (41) 


6 (35) 


17 (100) 


8 


24 


57-64 


19 (79) 


23 (96) 


24 (100) 


9 


20 


65-72 


13 (65) 


13 (65) 


20 (100) 


10 


18 


73-80 


6 (33) 


12 (67) 


18 (100) 


11 


21 


81-88 


6 (29) 


10 (48) 


21 (100) 


12 


24 


89-96 


4 (17) 


6 (25) 


24 (100) 


13 


13 


97-104 


2 (15) 


3 (23) 


13 (100) 


14 


10 


105-112 


1 (10) 


2 (20) 


10 (100) 


15 


10 


113-120 


2 (20) 


2 (20) 


10 (100) 



6 C. DALE BEERS 

The first dividing specimens of B. gracilis appeared in one of four urchins that 
had fed for 21 hours. Of 12 urchins that had fed for 22, 23, or 24 hours, B. gracilis 
was dividing in nine. Although B. gracilis is also a fairly large ciliate in which 
dividing specimens are easy to recognize, the number of such specimens was never 
great in any urchin of the entire study; usually it amounted to 2-A% of the total. 
Evidently division is resumed in B. gracilis soon after the arrival of fresh food 
in the rectum. 

Periods 46. The division of E. borcalis continued in many of the urchins 
(41-48%), but M. indomita was still not dividing. In Period 4, dividing speci- 
mens of B. gracilis were present in 81% of the urchins; in Periods 5 and 6, and 
indeed in all subsequent periods, it was dividing in all the urchins. 

In an ideal experiment the history of each of the ciliates should be followed 
in one and the same urchin by removing samples at intervals throughout the ex- 
periment. Unfortunately, this procedure is not feasible at present, and an urchin 
must be sacrificed at each examination. Such an examination may not reveal cor- 
rectly the actual physiological condition of the ciliates with respect to division. 
For example, reference to Table I (Period 4, Column 4) shows that E. borealis 
was dividing in 10 of the urchins ; it is understood that it was not dividing in the 
remaining 1 1 at the time of the examinations. But in some of the 1 1 , there 
was great variation in size among the individuals and they contained many 
gastrioles, suggesting that they had already divided at least once and were preparing 
for another division. Thus, in many urchins in which E, borealis was recorded as 
non-divisional, it was probably in a cycle of division and the percentages recorded 
in the table are actually low. These same comments apply to M. indomita, 
beginning with Period 7. 

Period 7. This period is of special interest, in that dividing specimens 
of M. indomita made their first appearance in many of the urchins (35%). 

The first dividing specimens of M. indomita appeared in a 49-mm. urchin that 
had fed for 50 hours, although they amounted to only \% of the total. The di- 
vision of M. indomita offered special difficulties of observation, since it is a 
slender, flattened, transparent ciliate, which is very active in samples of enteric 
fluid diluted with sea water for making counts. Magnifications high enough to 
reveal the condition of the macronucleus in living specimens (about 45 X ) render 
especially difficult the counting of specimens in samples. Therefore, the presence 
of a cytosomal constriction was used as the sole criterion of division and many 
pre-divisional specimens with elongated macronuclei were undoubtedly overlooked. 
The number of dividing individuals identified and recorded by the method never 
exceeded 5%. 

Period 8. In this period the division of E. borealis and M. indomita attained 
its maximal incidence in the urchins (79% and 96%, respectively). 

It is scarcely necessary to mention that population densities increased measur- 
ably in all the ciliates as division continued. 

Periods 9-15. In these periods the percentage of urchins that contained dividing 
specimens of E. borcalis and M. indomita gradually decreased, but with minor 
fluctuations which may have resulted from variations in the total numbers of 
urchins examined. 

In urchins kept four days or longer with food, it is somewhat difficult to relate 



CILIATES OF STRONGYLOCENTROTUS 7 

satisfactorily the division of E. borealis and M. indoinita to the amount of food 
in the gut, since many urchins stop feeding after three days and then feed irregu- 
larly. Sixteen urchins which had remained 9-14 days with food were examined. 
Two had dividing specimens of E. borealis and one had those of M. indoinita. Con- 
ditions in these urchins probably approximated those found in urchins under natural 
conditions ; that is, the amount of food in the gut was variable and the condition of 
the two ciliates was unpredictable, though usually non-divisional. 

DISCUSSION 

The results show conclusively that under certain experimental conditions (star- 
vation of the host, followed by generous feeding) there is a direct relation between 
the feeding of the urchin and the division of its ciliates : starved urchins contain 
no dividing ciliates, whereas many urchins fed on laminaria contain great numbers. 
With reference to the time at which division is resumed in experimentally fed 
urchins, the results permit these conclusions : division is resumed in B. gracilis 
after 20-30 hours of feeding; a period of division is initiated in E. borealis after 
10-15 hours of feeding; a similar period is initiated in M. indomita after 50-60 
hours. The evidence indicates that division continues indefinitely in B. gracilis, 
provided any appreciable amount of food remains in the rectum. The duration of 
the period of division of E. borealis and M. indomita is difficult to determine, owing 
to irregularities in the feeding habits of the host ; the evidence indicates that it 
continues for two to three days and then subsides in most of the urchins. 

It is clear that the results cast serious doubt on my earlier postulate to the effect 
that cycles of division are inherent in E. borealis and M. indomita, though they do 
not actually disprove the existence of such cycles. Nevertheless, it seems more 
likely that the occasional outbreaks of divisional activity are associated with periods 
of generous feeding on the part of the host. 

For the present the precise factors that are responsible for the division of the 
ciliates must remain unidentified. Presumably the presence of abundant food in 
the urchin gut results in a great increase in the numbers of bacteria that are both 
available and suitable as food for the ciliates. There is no doubt that the bacterial 
flora is greatly augmented by generous feeding on the part of the host. An examin- 
ation of enteric fluid from a well-fed urchin reveals great numbers of bacteria, 
whereas fluid from a starved urchin contains relatively few. It is recognized that 
absolute numbers may be of little significance, for some ciliates are extremely se- 
lective in their ingestion of bacterial food. For example, Kidder (1941, p. 471) 
isolated from a mass culture of Tillina canalijera 26 types of bacteria, only one of 
which was suitable to maintain the growth of the ciliate. Nevertheless, the many 
gastrioles in ciliates from well-fed urchins stand as proof of generous feeding, and 
the division of the ciliates has been conclusively demonstrated. Thus, it is clear 
that suitable bacteria were present in plentiful numbers, and it is assumed that 
division was the direct result of increased food ingestion. Whether the digestive 
juices of the urchin are important in stimulating division is unknown. 

Until the dietary requirements of the three ciliates are conclusively established, 
the facts seem to permit the following interpretation of the normal urchin-ciliate 
relationship. Ingested material is nearly always present in the rectum of any urchin 



C. DALE BEERS 

collected in nature, and this material suffices to support a bacterial flora that is 
both adequate and suitable to maintain B. gracilis in a constantly dividing state. 
The amount of ingested material in the stomach and intestine of such an urchin is 
demonstrably variable. The flora that it supports is considerable, but usually the 
flora suffices merely to maintain E. bore alls and M. indomita in a non-dividing 
state. Occasionally an urchin finds a plentiful supply of food and fills to re- 
pletion. Then the numbers of suitable bacteria increase to the extent that they are 
adequate to initiate and sustain a period of division. 

SUMMARY 

1. The study deals with division in the ciliates Entodiscus borealis, Madsenia 
indomita, and Biggaria gracilis. All the urchins collected in nature contained 
dividing specimens of B. gracilis, but only 3% of them contained those of E. borealis 
and M. indomita. 

2. Urchins were kept without food for 2-3 weeks ; in these, dividing specimens 
of the ciliates were absent. 

3. After about two weeks of starvation, the urchins were supplied with gener- 
ous amounts of the kelp Laminaria, and the percentage of the urchins that con- 
tained dividing specimens of each of the ciliates was recorded by successive 8-hour 
periods during 5 days of feeding. 

4. Division began in E. borealis after 10-15 hours of feeding by the host, and 
dividing individuals were present for about three days in 33-79% of the urchins. 

5. Division began in M. indomita after 50-60 hours, and dividing specimens 
were present for about two days in 25-96% of the urchins. 

6. Division began in B. gracilis after 20-30 hours, and dividing specimens were 
present in all the urchins after the second day. 

7. Division appears to continue indefinitely in B. gracilis, provided any appreci- 
able amount of food is present in the urchin gut. Although division appears to 
occur discontinuously in E. borealis and M. indomita, it is doubtful that cycles of 
division are inherent in them, as postulated earlier. It seems more likely that their 
division is correlated with the copious ingestion of suitable bacteria, whose numbers 
are greatly increased by the presence of abundant food in the urchin gut. 

LITERATURE CITED 

BEERS, C. D., 1948. The ciliates of Strongyloccntrotus drocbachicnsis: Incidence, distribution 

in the host, and division. Biol. Bull., 94: 99-112. 
BEERS, C. D., 1961. The obligate commensal ciliates of Strongyloccntrotus droebachiensis: 

Occurrence and division in urchins of diverse ages ; survival in sea water in relation to 

infectivity. Biol. Bull., 121: 69-81. 
KIDDER, G. W., 1941. The technique and significance of control in protozoan culture. In: 

Calkins and Summers : Protozoa in Biological Research, pp. 448-474. New York : 

Columbia University Press. 
POWERS, P. B. A., 1933a. Studies on the ciliates from sea urchins. I. General taxonomy. 

Biol. Bull., 65: 106-121. 
POWERS, P. B. A., 1933b. Studies on the ciliates from sea urchins. II. Entodiscus borealis 

(Hentschel), (Protozoa, Ciliata), behavior and morphology. Biol. Bull., 65: 122-136. 
SWAN, E. F., 1961. Some observations on the growth rate of sea urchins in the genus 

Strongylocentrotus. Biol. Bull, 120: 420-427. 



BIOLOGICALLY ACTIVE PEPTIDES FROM THE PERICARDIAL 
ORGANS OF THE CRAB CANCER BOREALIS x 

FRANK A. BELAMARICH 

Biological Laboratories, Harvard University, Cambridge 38, Massachusetts 2 

Pericardial organs are neurosecretory structures located in the pericardial 
cavity of decapod and stomatopod crustaceans. Alexandrowicz (1952, 1953) was 
the first to figure these structures and, with Carlisle, to show that low concentrations 
of aqueous extracts acted as a potent cardio-excitor (Alexandrowicz and Carlisle, 
1953). Extracts produced a marked increase in amplitude and frequency in all the 
crustacean species studied except Maia squincido, which exhibited an increase in 
amplitude, but a decrease in frequency. These investigators also found that blood 
collected from the pericardial cavity produced excitation similar to that of pericardial 
organ extracts. Cooke (unpublished data) has demonstrated that cardio-excitor 
hormone is released by electrical stimulation of the pericardial organs. 

Carlisle (1956) reported finding two active areas when pericardial organ ex- 
tracts were subjected to paper chromatography, both of which gave color reactions 
indicative of indole alkylamines. One of the active areas was destroyed by ortho- 
diphenol oxidase more rapidly than the other, an action that led Carlisle to believe 
the latter to be a precursor. Carlisle speculated that the active substance might be 
an ortho-dihydroxytryptamine, and later considered 5,6-dihydroxytryptamine the 
most likely possibility (Carlisle and Knowles, 1959). 

Maynard and Welsh (1959) found 5-hydroxytryptamine to be present in peri- 
cardial organ extracts, but at concentrations too low to account for the marked 
activity of the extract. These authors demonstrated that pericardial organ extracts 
could be inactivated by trypsin and chymotrypsin, and that the active principle was 
(1) dialyzable, (2) soluble in aqueous and alcoholic solvents, but only sparingly 
soluble in acetone, and (3) heat-stable at neutral and acid pH. They tentatively 
identified the neurosecretory material as a peptide. 

The purpose of this paper is to present further data from studies on the nature 
of the pericardial organ neurohormone. 

MATERIALS AND METHODS 

Cancer borealis and Cancer irroratus were used as a source of pericardial 
organs ; lobster hearts were used for assay. Animals were obtained from the Crab 
and Lobster Company, Boothbay Harbor, Me., or from local sources. Pericardial 
organs were removed by the method of Maynard and Welsh (1959) and either used 

1 This research was supported in part by a predoctoral fellowship from the National Insti- 
tutes of Health. 

2 Present address : Department of Biochemistry, School of Medicine, University of Buffalo, 
Buffalo 14, New York. 

9 



10 FRANK A. BELAMARICH 

immediately or freeze-dried and stored in a desiccator at 10 C. Pericardial organs 
stored by this method for more than one year were still active. 

Extracts of pericardial organs were made by grinding either the fresh or freeze- 
dried tissue in a glass homogenizer with the appropriate solvent. Aqueous extracts 
were heated at 100 C. for a few minutes and centrifuged, the sediment being 
washed several times and the washes pooled with the supernatant. For paper 
chromatography or paper electrophoresis, extracts were generally made in 95% 
ethanol and washed with petroleum ether. 

All extracts not made in Homarus perfusion fluid (Welsh and Smith, 1960, 
after Cole) were lyophilized or taken to dryness under reduced pressure. The 
residue was redissolved in perfusion fluid before assaying. Extracts were tested on 
a perfused isolated lobster heart which was kept at a constant temperature. Al- 
though individual experiments were done at constant temperatures, the range of 
temperatures during the course of these experiments was 15 to 20 C. A small 
volume (0.5 ml.) of the extract was introduced into the perfusion medium by 
syringe without disturbing the rate of flow of the fluid, and the action of the heart 
was recorded on a smoked drum. 

Paper electrophoresis of pericardial organ extracts was carried out utilizing 
Whatman 3 MM paper and buffers (acetate, phosphate, Tris, borate) with a pH 
range of 2.6 to 10.1 and ionic strength of 0.05. Voltage was usually 11.4 V./cm., 
although runs were also made at 5 V./cm. The extract, from 1-10 pairs of peri- 
cardial organs, was taken down to dryness under reduced pressure. It was then 
redissolved in a small amount (50-100 jul.) of 95% ethanol and streaked at the 
center of a prewashed paper. Lengths of run were 1 to 24 hours, after which 
papers were either air-dried, oven-dried at 60 C. and cut into sections, or cut 
while still wet. The latter method was the method of choice. The section which 
included the origin and one section to either side of the origin was of 1 cm. 
width, while all other sections were 2 cm. The sections were eluted into 3 ml. of 
solvent, freeze-dried if run in volatile buffers, and assayed. For staining, strips 
were cut from the long axis of the paper before sections were made. 

Chromatography of pericardial organ extracts was performed by the ascending 
one-dimensional method, utilizing n-butanol/acetic/water (60:15:25) and What- 
man 3 MM paper. Papers were pre-washed with distilled water and ethanol. De- 
velopment was usually 12-17 hours at room temperature. An extract from 4 to 20 
pairs of pericardial organs was applied as a streak and developed, after which a 
longitudinal strip was cut from the paper and sprayed with various staining 
reagents. In this way the area of biological activity was correlated with staining 
reaction. 

Eluates of the areas which produced activity and which were ninhydrin-positive 
were hydrolyzed by the method of Consden, Gordon and Martin (1947). The hy- 
drolysates were spotted on Whatman No. 1 paper, along with amino acid standards, 
developed in n-butanol/acetic/water (60:15:25) or n-butanol/pyridine/water 
(1:1:1). All hydrolysis experiments were done with pericardial organs from 
Cancer borealis. 

The confirmation of acidic and basic amino acid residues was accomplished 
with paper electrophoresis employing Schleicher and Schuell number 589 paper and 
0.05 M acetate buffer at pH 4.7. The hydrolysate, after being dried over NaOH, 



PERICARDIAL ORGAN ACTIVE PEPTIDES 



11 



was taken up in dilute NH 4 OH and applied to the center of the paper. A paper 
with lysine, glycine and glutamic acid applied at the origin was run simultaneously 
with the hydrolysate as a standard. 

RESULTS 

[nactivation of pericardial organ active suhstance was accomplished by incu- 
hating extracts made in Ho mar us perfusion fluid (pH 7.6) with crystalline trypsin 
or chymotrypsin (approximately 10 ^g./ml. extract) for one hour at 37 C. The 
enzymes alone had little, if any, effect on the heart. Heated enzyme incuhated with 
pericardial organ extract did not affect the activity. These results confirm those 
obtained by Maynard and Welsh (1959). Trypsin was used periodically to in- 
activate extracts and eluates from paper electrophoresis and chromatography. This 
gave assurance that the activity was due to pericardial organ cardio-excitor 
substance. 

A typical assay of a pericardial organ extract after electrophoresis is shown in 
Figure 1. As was true in most assays, the activity was spread over several centi- 
meters with a major peak at one fraction. Where activity was found to be equally 
distributed between fractions, it is assumed that the fraction boundary line cut the 
activity peak in two. 

The data in Figure 2 demonstrate that the isoelectric point of the active material 
is near pH 3, and an increase in pH above this point causes a corresponding increase 





FIGURE 1. A typical assay of pericardial organ extract after paper electrophoresis. 
Extract of four pairs of Cancer borcalis pericardial organs. Each fraction eluted into 3.0 ml. 
perfusion fluid and assayed on a perfused isolated lobster heart. Tris buffer, pH 7.7, 14 hours, 
150 V. 



12 



FRANK A. BELAMARICH 



b 



15 



10 






o 




FIGURE 2. Isoelectric point of pericardial organ cardio-excitor material. 

Buffer ionic strength 0.05. 



in migration toward the anode. A series of electrophoresis runs at pH 9.0, of 
increasing lengths of time, was made in an attempt to correlate staining reaction 
with migration of activity. Of the numerous reagents applied, only ninhydrin 
consistently stained the active area. 

Tests of eluates of chromatogrammed pericardial organ extracts showed activity 
from two areas. This activity resembled but did not equal that produced by crude 
extracts. Such activity could be inactivated by trypsin. The activity was localized 
to two ninhydrin-positive areas (Fig. 3). 

Chromatography of acid hydrolysates of the two active areas showed that 
these were two peptides of similar nature. These peptides have been given the 
designation peptides A and B. Hydrolysates of these peptides were subjected to 
paper chromatography and paper electrophoresis under similar conditions. These 
revealed the presence of glutamic acid and a positively charged amino acid (lysine 
or arginine), as well as the presence of neutral amino acids. Based on staining 
intensity, the glutamic acid was present in greater concentration than the positively 
charged amino acid. One other acidic residue, which had a migration rate slightly 
higher than that of glutamic acid, was also present in low concentration (Fig. 4). 
The neutral amino acids were categorized by their R f values and comparison with 
standards in two chromatographic solvents, and were tentatively identified as serine 
in both peptides, tyrosine in peptide A, and methionine in peptide B. Each peptide 
contained one other residue which appeared in the region of valine or phenylalanine. 

In order to demonstrate that one was dealing with the same material in electro- 
phoresis and chromatography, a pericardial organ extract was subjected to el- 
ectrophoresis and the area of activity determined by assay. A matching area from 
a simultaneously run paper was then eluted and subjected to paper chromatography. 
Examination showed that the two active, ninhydrin-positive areas present in the 



PERICARDIAL ORGAN ACTIVE PEPTIDES 



13 










-R f 05 




FIGURE 3. Chromatography and assay of an extract of 20 pairs of pericardial organs ; 
n-butanol/acetic/water ; 14 hours. Ninhydrin stain. A and B are active areas, C is control. 
Arrow marks point at which eluate reaches heart. 

chromatogrammed pericardial organ extract were also present in eluates of active 
areas following electrophoresis. 

DISCUSSION 

Maynard and Maynard (1962) estimate that the neurosecretory granules may 
form 2% or less of the volume of the trunks of hrachyuran pericardial organs. 
Since the average dry weight of freeze-dried pericardial organs from Cancer borealis 
was approximately 0.6 mg. /animal, the amount of material in the neurosecretory 



14 



FRANK A. BELAMARICH 



TTTTTI 



Illlllin 







Origin 



Glutamic 



Glycine 



Lysine 



Hydrolysate of 
Peptide A 



Standards 



FIGURE 4. Electrophoresis of hydrolysate of peptide A. Acetate buffer, 
pH 4.7, ionic strength 0.05, 200 V., 4 hours. 

granules can be calculated to be approximately 12 /xg. /animal. Although there 
is disagreement as to whether neurohormones are attached to a carrier protein or 
whether they are split off from a parent molecule (Bern, 1962), large protein 
molecules appear to be present in most, if not all, neurosecretory granules. The 
presence of such a protein, along with other components, would make the figure 
for actual neurohormone material present in the pericardial organ somewhat less 
than the calculated 12 //.g. /animal. 

Pericardial organ extracts subjected to paper electrophoresis and paper chroma- 
tography apparently lose some activity. Smyth (1959 ) also noted this with crayfish 
pericardial organs, and found that developing chromatograms in an atmosphere of 
nitrogen or in the presence of ascorbic acid did not prevent loss of activity. With 
paper electrophoresis there was a correlation between the length of run and activity 
recovered, i.e., short runs retained proportionally more activity. 

The fact that there are two active components is not surprising. Carlisle (1956) 
reported finding two areas of activity after chromatography of pericardial organ 
extract. Of interest in this respect is the report by Alexandrowicz (1953) in which 



PERICARDIAL ORGAN ACTIVE PEPTIDES 15 

he distinguishes two types of neurosecretory fibers contributing to the pericardia! 
organ system of Squilla. Electron microscopy (Knowles, 1960) shows two types 
of neurosecretory granules in pericardial organs of Squilla, differing in size and 
internal structure. More recently, Maynard (1961) demonstrated the presence 
of three types of secretory cells contributing fibers to the pericardial organ-anterior 
ramification complex in several species of decapod crustaceans. When examined 
by electron microscopy, Cancer pericardial organs showed neurosecretory granules 
of only one size, but the other brachyuran genera examined contained granules of 
two or three size groups (Maynard and Maynard, 1962). 

The data indicate that the active compounds extracted from pericardial organs 
are acidic peptides with an isoelectric point near pH 3. The presence of propor- 
tionally more glutaminyl residues than positively charged residue (lysine or 
arginine) fits well with this conclusion. The proposed amino acid composition of 
the peptides fits the requirements for enzymatic degradation by trypsin and chymo- 
trypsin, since trypsin is specific for lysine or arginine linkages, while chymotrypsin 
is specific for aromatic, as well as leucyl, methionyl, arginyl, and glutaminyl bonds. 

The identification of the neutral and basic amino acid residues has been made 
on a tentative basis. An examination of peptides A and B at this time indicates the 
possibility of a single amino acid difference. A small difference, especially in the 
neutral amino acids, would account for the inability to separate the peptides by 
means of paper electrophoresis. The residue exhibiting slightly higher negative 
charge than glutamic acid has not been identified. It is possible that a non-amino 
acid residue is present with the peptide. Smyth (1959) noted that after chromatog- 
raphy of crayfish pericardial organs in methanol/water/pyridine the active area 
stained lightly with ammoniacal silver nitrate as well as with ninhydrin. 

Tryptophan analysis was not feasible because of the small amount of active ma- 
terial present in pericardial organs. However, it should be pointed out that peri- 
cardial organ extracts and 5-hydroxytryptamine have a qualitatively similar action 
on the decapod heart (Maynard and Welsh, 1959; Cooke, 1962), and this 
similarity may be due to the presence of tryptophan in the peptide. 

Carlisle has considered 5,6-dihydroxytryptamine as the most likely possibility 
for the pericardial organ neurohormone. Such a compound has been shown to be 
extremely labile, exhibiting spectral shifts after 30 minutes in aqueous solution at 
pH 7, and after 5 minutes at pH 8 (Schlossberger and Kuch, 1960). There have 
been no published data to show that such a compound, in fact, exists in pericardial 
organs of decapod crustaceans, although the possibility is not discounted. If pres- 
ent, it may be acting at a different site on the cardiac ganglion, as has been demon- 
strated for 5-hydroxytryptamine (Cooke, 1962). The rate at which pericardial 
organ extracts used in this study lose activity, as well as other criteria, indicates 
that the activity of these extracts is not due entirely to 5,6-dihydroxytryptamine. 

It is interesting to note that enzyme inactivation studies of crustacean chroma- 
tophorotropins and retinal pigment hormones show that all these neurosecretory 
products are inactivated by either trypsin or chymotrypsin or both of these enzymes 
(Knowles, Carlisle and Dupont-Raabe, 1956; Perez-Gonzalez. 1957; Kleinholz, 
Esper, Johnson and Kimball, 1961). It is well known that the neurosecretory 
hormones of the vertebrate hypothalamic-neurohypophyseal system are peptides. 
Future research may show that one of the main functions of neurosecretory systems 
is the production of biologically active peptides. 



16 FRANK A. BELAMARICH 

I wish to express my appreciation to Professor John H. Welsh for his con- 
tinued guidance and interest in this problem, to Dr. Russell F. Doolittle for his 
invaluable advice, and to Misses Carolyn Sharpe and Joyce Zipf for help in the 
dissections. 

SUMMARY 

1. The isoelectric point of pericardial organ cardio-excitor material, determined 
by paper electrophoresis, is near pH 3. 

2. Paper chromatography demonstrates two active compounds in pericardial 
organ extracts which stain with ninhydrin and upon acid hydrolysis are identified 
as peptides. These peptides differ in composition by only a small number of 
residues. 

3. The peptides contain proportionally more glutaminyl residues than positively 
charged residue (lysine or arginine), one other negatively charged residue, and a 
small number of neutral amino acids. 

LITERATURE CITED 

ALEXANDROWICZ, J. S., 1952. Notes on the nervous system in the stomatopods. I. The system 

of median connectives. Pubbl. Stas. Zool. Napoli, 23: 201-214. 
ALEXANDROWICZ, J. S., 1953. Nervous organs in the pericardial cavity of the decapod Crustacea. 

/. Marine Biol. Assoc., 31 : 563-580. 
ALEXANDROWICZ, J. S., AND D. B. CARLISLE, 1953. Some experiments on the function of the 

pericardial organs in Crustacea. /. Marine Biol. Assoc., 32: 175-192. 
BERN, H. A., 1962. The properties of neurosecretory cells. Gen. Comp. Endocrin., Supplement, 

1: 117-132. 
CARLISLE, D. B., 1956. An indole-alkylamine regulating heart-beat in Crustacea. Biochem. J., 

63: 32-33 P. 

CARLISLE, D. B., AND F. KNOWLES, 1959. Endocrine Control in Crustaceans. Cambridge Uni- 
versity Press, Cambridge. 
CONSDEN, R., A. H. GORDON AND A. J. P. MARTIN, 1947. The identification of lower peptides in 

complex mixtures. Biochem. J., 41 : 590-596. 
COOKE, I., 1962. Effects of the pericardial organ neurosecretory substance on the crustacean 

heart. Gen. Comp. Endocrin., 2 : 29. 
KLEINHOLZ, L. H., H. ESPER, C. JOHNSON AND F. KIMBALL, 1961. Characterization and partial 

purification of crustacean eyestalk hormones. Anier. Zool., 1: 366. 
KNOWLES, F. G. W., 1960. A highly organized structure within a neurosecretory vesicle. 

Nature, 185: 709-710. 
KNOWLES, F. G. W., D. B. CARLISLE AND M. DUPONT-RAABE, 1956. Inactivation enzymatique 

d'une substance chromactive des insectes et crustaces. C. R. Acad. Sci., Paris, 242: 825. 
MAYNARD, D. M., 1961. Thoracic neurosecretory structures in Brachyura. II. Secretory 

neurons. Gen. Comp. Endocrin., 1 : 237-263. 
MAYNARD, D. M., AND E. A. MAYNARD, 1962. Thoracic neurosecretory structures in Brachyura. 

III. Microanatomy of peripheral structures. Gen. Comp. Endocrin., 2: 12-28. 
MAYNARD, D. M., AND J. H. WELSH, 1959. Neurohormones of the pericardial organs of 

brachyuran Crustacea. /. Physiol., 149: 215-227. 
PEREZ-GONZALEZ, M. D., 1957. Evidence for hormone-containing granules in sinus glands of 

the fiddler crab Uca pugilator. Biol. Bull., 113: 426-441. 
SCHLOSSBERGER, H. G., AND H. KUCH, 1960. Synthese des 5,6-Dihydroxy-tryptamins. Chciu. 

Ber., 93: 1318-1323. 
SMYTH, T., JR., 1959. Paper chromatography of pericardial organs of crayfish. Biol. Bull., 

117: 426. 
WELSH, J. H., AND R. I. SMITH, 1960. Laboratory Exercises in Invertebrate Physiology. 

Revised Edition. Burgess Publ. Co., Minneapolis. 



THE GENETICS OF ARTEMIA SALINA. II. WHITE EYE. 
A SEX-LINKED MUTATION 1 

SARANE THOMPSON BOWEN 

Department of Biology, San Francisco State College, 
San Francisco 27, California 

Although there have been many cytological studies of sex determination in the 
Crustacea (reviewed by Niiyama, 1959), this paper will describe the first genetic 
study of crustacean sex determination. It will be shown that the female is the 
heterogametic sex in the brine shrimp. Sex differentiation in Arteinia has been 
discussed by Bowen and Hanson (1962). 

The culture medium and the genetic techniques used in the following experi- 
ments were described in the first paper in this series (Bowen, 1962). 



The author would like to thank Miss Jean Hanson, Mr. Dan Straus, and Mr. 
H. Stuart Williamson for their assistance throughout this study. 

MATERIALS 

Origin of stock #5 

In all the races of Arteinia, the wild-type eye is black. The recessive autosomal 
gene, r, for red eyes arose as a spontaneous mutation in the Utah race. It was 
described in the first paper in this series. A second autosomal gene, c, which de- 
termines "crinkle eyes," has not been described previously. This mutation arose 
spontaneously in the California race (San Francisco Bay) and was discovered by 
Miss Jean Hanson in 1960. At the age of three to five weeks, shrimp of the cc 
genotype develop an extra patch of pigment cells on the eyestalk. The two muta- 
tions have been combined in a single line, designated as stock #5 in our laboratory. 
Young shrimp in this rr cc stock have red eyes. At sexual maturity their eyes 
turn brown or black. At the age of five weeks the pigment in the normal eye field 
is black, but the pigment in the "crinkles" patch on the eyestalk is still red. 

Origin of the white eye mutation 

A \vild-type male from a cyst collected at Great Salt Lake, Utah, was mated to 
a female of the #5 stock. An Fj male with the RrCc genotype was backcrossed to 
another #5 stock female. In this progeny the author found one white-eyed male 
in December of 1961. 

1 This research was supported hy a grant from the National Science Foundation (NSF 
G-23863). 

17 



18 



SARANE THOMPSON BOWEN 
RESULTS 



The terms pigmented or wild-type phenotype are used below to designate non- 
white-eyed shrimp. Because the first white-eyed male had the genotype Rr, some 
of the pigmented shrimp were red-eyed and some were black-eyed in Experiments 
A, B, and C. 

Experiment A 

The first white male was mated to a female from the #5 stock. This mating 
is represented as Generation I in Figure 1. Three broods of pigmented F x nauplii 
were produced : a total of 14 males and 19 females lived to maturity. The F I 
shrimps were mated to each other to produce an F 2 which consisted of 14 white 
males, 16 pigmented males, and 32 pigmented females (shown in Generation III of 
the pedigree). In this first experiment, the mode of inheritance of white eye in 
Artemia resembled that of white eye in Drosophila. 



I 





BACKCROSS 



14 



i 



19 






4 16 32 



17 



16 



V 




OUTCROSS 



164 



281 




114 



143 



\l 



tot 




60 



67 




54 



99 



I 



154 



69 



6 



58 



FIGURE 1. Pedigree showing the distribution of white eye in the first six generations. 
Conventional genetic notation is used : solid symbols indicate shrimps with mutant phenotype ; 
open symbols indicate shrimps with wild-type phenotype (pigmented eyes). Squares represent 
males ; circles represent females. The number written under each symbol indicates the number 
of progeny in that class. 



SEX LINKAGE IN ARTEMIA 19 

Experiment B 

An F! female from Experiment A was backcrossed to the first white male. 
Surprisingly, the progeny consisted of 17 white males and 16 pigmented females 
(shown in Generation III of the pedigree in Figure 1). These shrimp were mated 
inter se and produced 164 white males and 281 pigmented females. These were 
mated to each other and the progeny consisted of 114 white males and 143 pig- 
mented females. Thus, the mating of a white male to the pigmented daughter of 
a white male results in a pure-breeding stock of white males and pigmented females. 
This stock is carried in our laboratory as stock #9. 

Experiment C 

Three females from stock #9 were outcrossed to wild-type males which had 
hatched from cysts collected in Queraado, New Mexico (U.S.A.). The F t 
shrimps (shown in Generation V of the pedigree) were all pigmented: 60 males 
and 67 females. Six of the F x females were mated to #9 stock white males, and 
again the progeny were all pigmented : 69 males and 58 females. Four of the F x 
males were mated to females from the #9 stock. These four matings produced 
54 white males, 99 pigmented males, 154 pigmented females, and one white female 
(shown in the sixth generation of the pedigree). This first white female appeared 
in May of 1962. She was mated to a white brother and produced all white-eyed 
progeny. 

THE HYPOTHESIS OF PARTIAL SEX LINKAGE 

The experimental results can be accounted for by the following assumptions : 
(1) The mutant gene w which determines white eyes is recessive to its wild-type 
allele W. (2) The females are heterogametic. The chromosome constitution of 
the female will be represented as XY; the males will be XX. (3) The white 
locus is partially sex-linked. Because it is on the homologous segment of the sex 
chromosomes, both males and females may be WW , Ww, or ww. 

The first white female probably arose as the result of a crossover between the 
white locus and the "sex locus." It is not known whether sex is determined by 
two alleles at one true sex locus or by several loci on the differential segment of the 
X or the Y chromosome. If- there were several loci governing sex determination, 
the "sex locus" would be designated as the place where the differential segment 
of a sex chromosome joins the homologous segment (Fig. 2). Both concepts of 
a sex locus would be in accord with the experimental results. 

The cytological studies of Artemia (reviewed by Barigozzi, 1957) have not 
revealed the presence of a pair of chromosomes of unequal length. For this reason 
the X and Y chromosomes in Figure 2 are shown to be the same length. The 
length of the differential segment in relation to the homologous segment is an 
arbitrary choice. Only one differential segment is shown on each sex chromosome 
although in many species two such segments have been found (reviewed by 
Darlington, 1958). 

Partial sex-linkage has been reported in three genera of viviparous killifishes 
(family Poeciliidae) : in Aplocheilus latipes (Aida, 1921), in the guppy, Lebistes 
reticulatus (Winge and Ditlevsen, 1947) and in the platyfish, Xiphophorus (Platy- 



20 



SARANE THOMPSON BOWEN 



DIFFERENTIAL 
SEGMENT 

HOMOLOGOUS 
SEGMENT 




SEX LOCUS 
WHITE LOCUS 



Y X 



FIGURE 2. Diagram of the sex chromosomes of Artemia. Crossing over may occur 
between the white locus and the sex locus. 

poecilus) macidatus (Gordon, 1937, 1947; reviewed by Bellamy and Queal, 1950). 
Crossover of a gene from the X to the Y chromosome has also been reported at 
the bobbed locus of Drosophila uielanogaster (Stern, 1929; Neuhaus, 1937). Al- 
though many characteristics in the human have been reported to be partially sex- 
linked, Morton (1957) has found that in every instance the data fell short of 
statistical significance. 

EXPLANATION OF THE EXPERIMENTAL RESULTS 

In Figure 3 the hypothesis of partial sex-linkage is applied to two experiments. 
On the left is a pedigree which describes Experiment A. If a white male is mated 




white 






wild 
(homozygous) 



white 




wild (F,) 



wild 





wild 



white 






wild 




white 



wild 



wild 



wild 



white 



wild 



FIGURE 3. Sex-linked inheritance in Artemia. Squares represent males ; circles represent 
females. The phenotype of each class is written beneath the symbol. The X and Y chromosomes 
are shown as in Figure 2. Crossing over between the X and Y does not occur in these two 
experiments. 



SEX LINKAGE IN ARTEMIA 



21 



to a homozygous wild-type (pigmented) female, all the F t progeny will be wild- 
type. If the F! shrimps are mated inter se, the F 2 phenotypic ratio would be : % 
white males :% wild-type males :% wild-type females. The observed ratio was 
14:16:32. 

On the right side of Figure 3 is a pedigree which explains the data in Experi- 
ment B. If a white male is mated to the heterozygous daughter of a white male, a 
pure-breeding stock of white males and phenotypically wild-type females will be 
established. That there were fewer males than females classified in Experiment B 
is probably due to the lower viability of the white phenotype (discussed below). 

Partial sex-linkage was proposed as the mode of inheritance of white eye before 
the fourth generation was classified. At that time it was predicted that a white- 
eyed female would be produced as the result of crossing over. Experiment C was 
designed to test the hypothesis of partial sex-linkage ; the results are in accord 
with the hypothesis. 

SUMMARY OF GENETIC DATA 

The data from Experiments A, B, and C, and from additional matings made in 
the summer of 1962 are combined in Table I. The number of broods may exceed 
the number of matings because some matings produced more than one brood. The 
viability of the white phenotype calculated from the progeny of matings of X W Y W 
females X X W X W males is 623/877 or 0.71. Good agreement is found when the 
viability is calculated from the all-male progeny of matings of X W Y W females 
X X W X W males: 99/152 or 0.65. The author is unable to explain the significant 
deviation from the expected ratio in the progeny of matings of X W Y W females 
X X W X W males. One would expect the ratio of pigmented females to pigmented 
males to be 2. The observed ratio is 247/152 or 1.6. 

The X and Y chromosomes carried by stock #9 are of special interest. The X 
chromosome is marked by the mutant gene w and was carried by the first white- 
eyed shrimp (an X W X W male). The mutant gene arose spontaneously in the X 
chromosome which evidently came from the #5 stock. The Y carries the wild 



Segregation of the gene w. 



TABLE I 

(In the genotype oj the female parent, the gene on the Y 
chromosome is underlined) 









Number of offspring 


Number of 


Number of 


Mating 
female X male 












matings 


broods 


(XV) (XX) 


White 
male 


White 
female 


Pigmented 

male 


Pigmented 
female 


Total 


25 


32 


WW X ww 








360 


362 


722 


61 


82 


wW X ww 


623 








877 


1500 


3 


4 


wVV X WW 








60 


67 


127 


14 


21 


wW X Ww 


99 


1 


152 


247 


499 


23 


31 


ww X ww 


166 


174 








340 



SARANE THOMPSON BOWEN 

allele and was present in the #5 female (X W Y W ) to which this male was mated in 
Generation I. Crossing over between this X w chromosome and this Y w chromo- 
some is rare. For example from the #9 stock matings (X W Y W females X X W X W 
males), a total of 1500 progeny consisted of white males and pigmented females. 
That is, no crossovers were detected among the 1500 female gametes tested. How- 
ever, one recombinant was found among the 499 offspring of matings of #9 stock 
females to X W X W males (shown in the fourth line in Table I). Since only one- 
fourth of the recombinant progeny of such matings can be detected, the crossover 
rate here is 4/499 or 1/125. The crossover frequency calculated from all the data 
from #9 females is 1/(125 + 1500) which is 1/1625 or 0.06%. 

The racial origin of these two chromosomes is unknown because they originated 
from the #5 stock which is derived from both the Utah and California races. In a 
later paper it will be shown that when a female is carrying the same X w chromo- 
some and a Y w from another source, the frequency of crossing over may be as high 
as 9%. Evidently a crossover suppressor mechanism is present in stock #9. 

SUMMARY 

1. This paper describes the first sex-linked gene discovered among the Crus- 
tacea. The experimental results can be accounted for by the following hypotheses : 
(1) the mutant gene w, which determines white eyes, is recessive to its wild allele 
W. (2) Female brine shrimp are heterogametic. The chromosome constitution 
of the female will be represented as XY; the males are XX. (3) The white locus 
is partially sex-linked. Because it lies on the homologous segments of the sex 
chromosomes, both males and females may have the genotype of WW ', Ww, or ww. 

2. In matings of Ww females to ww males, crossing over between the white 
locus and the sex locus may be detected. In the #9 stock, females have the geno- 
type X W Y W and the frequency of crossing over is low (1/1625 or 0.06%). This 
results in a "mother-to-daughter" inheritance of the W gene in the #9 stock. 

LITERATURE CITED 

AIDA, T., 1921. On the inheritance of color in a fresh-water fish Aplochciliis latipcs Temmick 

and Schlegel, with special reference to sex-linked inheritance. Genetics, 6: 554-573. 
BARIGOZZI, C., 1957. Differenciation des genotypes et distribution geographique d'Artcmia salina 

Leach: donnees et problemes. Annee Biol., 33: 241-250. 
BELLAMY, A. W., AND M. L. QUEAL, 1950. Heterosomal inheritance and sex determination in 

Platypoccilus macidatus. Genetics, 36: 93-107. 
BOWEN, S. T., 1962. The genetics of Artcmla salina. I. The reproductive cycle. Biol. Bull., 

122: 25-32. 
BOWEN, S. T., AND J. HANSON, 1962. A gynandromorph of the brine shrimp, Artemia salina. 

Genetics, 47: 277-280. 

DARLINGTON, C. D., 1958. The Evolution of Genetic Systems. Basic Books, Inc., New York. 
GORDON, M., 1937. Genetics of Platypoccilus. III. Inheritance of sex and crossing over of the 

sex chromosomes in the Platyfish. Genetics, 22: 376-392. 
GORDON, M., 1947. Genetics of Platypoecilus macnlatus. IV. The sex determining mechanism 

in two wild populations of the Mexican platyfish. Genetics, 32: 8-17. 
MORTON, N. E., 1957. Further scoring types in sequential linkage tests, with a critical review 

of autosomal and partial sex linkage in man. Amer. J. Hum. Genet., 9: 55-75. 



SEX LINKAGE IN ARTEMIA 



23 



NEUHAUS, M. H., 1937. Additional data on crossing over between X and Y chromosomes in 

Drosophila melanogaster. Genetics, 22: 333-339. 
NIIYAMA, H., 1959. A comparative study of the chromosomes in decapods, isopods and 

amphipods, with some remarks on cytotaxonomy and sex-determination in the 

Crustacea. Memoirs of the Faculty of Fisheries, (Hokkaido University), 7: 1-60. 
STERN, C, 1929. Untersuchungen uber Aberrationen des Y-chromosoms von Drosophila 

melanogaster. Zeitschr. induktive Abstammnngs- und Vererbungslehre, 51: 253-353. 
WINGE, O., AND E. DITLEVSEN, 1947. Colour inheritance and sex determination in Lebistes. 

Heredity, 1: 67-83. 




CHROMATOPHORE CONTROL AND NEUROSECRETION IN THE 
MUD SHRIMP, UPOGEBIA AFFINIS l 

MILTON FINGERMAN AND CHITARU OGURO 

Department of Zoology, Neivcomb College, Tulane University, Nezv Orleans, Louisiana, 

Akkeshi Marine Biological Station, Akkeshi, Japan, and Marine 

Biological Laboratory, Woods Hole, Massachusetts 

The control of chromatophores in the mud shrimp, Upogebia affinis, has not been 
described. This organism has been used in very few investigations of chro- 
matophores, and in these only as a source of tissues suspected of containing 
chromatophorotropins. Furthermore, only one account (Hanstrom, 1937) of the 
morphology of neuroendocrine structures in this organism is available. In 1948, 
Hanstrom summarized the results of his 20-year study of the supraesophageal 
ganglia and incretory organs in the Malacostraca. 

Systematists differ as to whether the tribe Thalassinidea, to which Upogebia 
belongs, should be placed with the Anomura or the Macrura. In 1960, Waterman 
and Chace included this tribe with the Macrura. On the other hand, in 1961, Green 
classified the Thalassinidea among the Anomura. 

The optic ganglia and sinus glands of the mole crab, Emerita talpoida, and of 
Upogebia are closely associated with the supraesophageal ganglia instead of oc- 
curring in the eyestalks as is the case in most decapods studied. According to 
Hanstrom (1937) the retinal structures and their associated nerve tracts are the 
only major components of the nervous system present in the eyestalks of these two 
species. In contrast, the eyestalk of the anomuran, Pagarus pollicaris, a hermit 
crab, contains the usual ganglia, such as the medulla terminalis, and the sinus gland, 
in addition to the visual structures (Hanstrom, 1937). 

Perkins and Kropp (1932) noted that eyestalks of Pa gurus longicarpus blanched 
the shrimp, Crangon boreas, which is not an anomuran. In contrast, no response 
was observed when extracts of eyestalks from Upogebia and Emerita were injected 
into the fiddler crab, Uca, by Carlson (1936) or into the prawn, Palaemonetes, by 
Hanstrom (1937). But Hanstrom did find that eyestalks of Pagurus pollicaris 
blanched eyestalkless Palaemonetes. Head extracts of Upogebia, on the other hand, 
darkened eyestalkless Uca (Carlson, 1936) and blanched eyestalkless Palaemonetes 
(Hanstrom, 1937). However, the question remained whether the effect of the 
head extracts was due to the sinus glands, the supraesophageal ganglia, or both 
structures, until Sandeen and Baldwin (1962) assayed glands and supraesophageal 
ganglia of Upogebia on Uca. Both organs yielded extracts that caused even more 
melanin dispersion than did extracts of the same tissues from Uca when assayed 
on Uca. 

In the meantime, Brown and Scudamore (1940) had postulated the presence of 

i This investigation was supported in part by Grant No. B-838 from the National Institutes 
of Health. 

24 



CHROMATOPHORE CONTROL IN UPOGEBIA 25 

at least two chromatophorotropins in the sinus gland of Pagurus pollicaris. Brown 
and Saigh (1946) found two antagonistic chromatophorotropins in central nervous 
organs of Upogebia affinis, Emerita talpoida, Pagurus pollicaris, and P. longicarpus. 
Because the assays were performed on Crangon, the antagonists in the central nerv- 
ous organs were termed Crangon body-lightening hormone (CBLH) and Crangon 
darkening hormone (CDH). 

Through the efforts of several investigators working in the early 1950's (e.g., 
Passano, 1951 ; Bliss and Welsh, 1952), the sinus gland was shown to be a storage 
and release organ, i.e., a neurohemal organ, rather than an actual site of hormone 
production. Chromatophorotropins appear to be products of neurosecretory cells, 
transported by axoplasmic flow from the site of formation to neurohemal organs. 
Miyawaki (1960) is the only investigator who described the cytology of neuro- 
secretory cells in anomurans. He found three types of such cells in the central 
nervous organs of the crabs Eupagurus ochotensis and Paralithodes brcvipes. 

The literature concerning (a) the morphology of neuroendocrine organs and 
(b) the physiology of chromatophores in typical Macrura is very extensive. In such 
forms the sinus gland typically resides in the eyestalk (Hanstrom, 1937). The 
chromatophore system of Macrura is highly evolved, pigment-concentrating and 
pigment-dispersing principles having been demonstrated (Fingerman and Aoto, 
1962). 

The general object of this investigation was to learn the origins and actions 
of chromatophorotropins in Upogebia. The specific aims were to determine (1) 
the responses of Upogebia to extracts of its own eyestalks, sinus glands, and supra- 
esophageal ganglia, and (2) the distribution of neurosecretory cells in the head 
of Upogebia. 

MATERIALS AND METHODS 

We are indebted to the personnel from the Supply Department of the Marine 
Biological Laboratory at Woods Hole, Massachusetts, and to Dr. Muriel I. Sandeen 
of the Duke Marine Laboratory at Beaufort, North Carolina, for furnishing speci- 
mens of Upogebia affinis. Specimens used in the bioassays were maintained in the 
laboratory at Woods Hole in aquaria supplied with constantly flowing sea water. 

Red chromatophores on the dorsal surface of the telson and uropods were 
staged according to the system of Hogben and Slome (1931). Stage 1 represents 
maximal concentration of the pigment, stage 5 maximal dispersion, and stages 2, 3, 
and 4 the intermediate conditions. Student's t test was used in the statistical 
analysis of the data. 

Tissue extracts were prepared by grinding the appropriate number of organs in 
sea water. The concentration was one-third of the organ complement from one 
mud shrimp per dose, 0.05 ml. Therefore, each dose contained either both eye- 
stalks, both sinus glands, or the supraesophageal ganglia from one mud shrimp. 

Paraffin sections of the eyestalks, sinus glands, and supraesophageal ganglia 
were prepared in the usual fashion. These structures were fixed in (1) Bouin's 
solution or (2) Helly's solution. Sections 8 and 10 /* thick were stained with (1) 
Mallory's trichrome, (2) Heidenhain's azan, (3) Gomori's chrome alum hema- 
toxylin-phloxin, or (4) aldehyde fuchsin. 



26 



MILTON FINGERMAN AND CH1TARU OGURO 



OBSERVATIONS AND RESULTS 
Responses of erythrophores in Upogcbia to tissue extracts 

The aim of this experiment was to observe the effects of eyestalks, sinus glands, 
and supraesophageal ganglia from Upogebia on eyestalkless specimens of Upogebia. 
The results are presented in Figure 1 where the results from three experiments 
(3, 3, and 4 test animals, respectively) are averaged. 

Aside from the responses to the extracts, inspection of Figure 1 reveals that 
the red pigment of eyestalkless Upogebia was maximally dispersed before injection 
of the extracts and sea water. Injection of the tissue extracts caused statistically 
significant degrees of pigment concentration, p < 0.001 for each extract, despite the 
fact that control injections of sea water evoked some pigment concentration. The 
chromatophore indexes measured 15 minutes after injection of extracts were used 
for the statistical calculation. Responses of the chromatophores were highest to 
extracts of supraesophageal ganglia and least to eyestalk extracts. 

Neurosecretory cells in Upogebia 

In view of the observation (Fig. 1) that extracts of the supraesophageal ganglia, 
sinus glands, and eyestalks of Upogebia caused a significant concentration of the 
red pigment in this organism, investigation of these structures for signs of neuro- 



4 
f 



ui 
<r 



Q. 
O 



2 
O 
a: 




024 

HOURS 

FIGURE 1. Relationships between chromatophore stage and time following injection of 
tissue extracts into eyestalkless Upogcbia. Circles, eyestalks ; dots, supraesophageal ganglia ; 
circles half filled on left, sinus gland ; circles half filled on bottom, control. 



CHROMATOPHORE CONTROL IN UPOGEB1A 



E- 



27 




FIGURE 2. Diagram of the eyestalk of Upogebia. E, eye; ON, optic nerve; SC, secretory cells. 



secretory activity seemed appropriate. Examination of sectional eyestalks revealed 
a group of 500-700 ovate cells whose cytology and staining properties are typical 
of neurosecretory cells. This cluster of cells lies in the dorsal half of the distal 
third of the eyestalk, partially surrounding the optic nerve, and is enclosed by a 
connective tissue sheath (Fig. 2). The cells are fairly uniform in size, averaging 
14 p. wide and 19 p. long, with an axon emerging from one end. The nucleus, which 
has a conspicuous nucleolus, is centrally located in some of the cells, eccentrically 
in others. The cytoplasmic granules appear to be neurosecretory products. For 
example, they stain pink with Heidenhain's azan and are positive to aldehyde 
fuchsin. In addition to the granules, some of the cells have cytoplasmic vacuoles. 
The cells are arranged in clusters resembling rosettes (Fig. 3). Because of the 
unique architecture of this group of cells, the structure will be referred to as the 
Rosette Body. Blood sinuses were noted in both the central and peripheral portions 
of the Rosette Body. Pores in the connective tissue sheath afford the blood ready 
access to these sinuses. 

Axons of the optic nerves contained granules that appeared to be neurosecretory 
products. These granules probably originated in the Rosette Body, inasmuch as the 
Rosette Body was the only structure observed in the eyestalk that could be the 



28 MILTON FINGERMAN AND CHITARU OGURO 




FIGURE 3. Detailed diagram of the cell arrangement in a portion of the Rosette Body. 

source of the secretory granules in the optic nerve. The axons from the Rosette 
Body appear to unite with the optic nerve just proximal to the basement membrane 
of the eye. Axons of the optic nerve could not be followed after they entered the 
supraesophageal ganglia. 

The general morphology of the sinus glands and supraesophageal ganglia con- 
forms to the description and photograph presented by Hanstrom (1937). Inspec- 
tion of histological sections of the supraesophageal ganglia revealed the presence 
of three distinct types of cells that had neurosecretory staining properties. The 
largest type has a cell body averaging 24 p. wide and 40 p. long and is restricted to 
the medullae terminales. The left and right medullae terminales are fused in the 
midline of the supraesophageal ganglia, as described by Hanstrom (1937). About 
six cells of this type occur. They have a round nucleus in the center of the cell 
body. A conspicuous nucleolus lies near the nuclear membrane. Near the periphery 
of the cytoplasm occur several vacuoles. The second type, similar in size and 
shape to the cells that compose the Rosette Body, possesses a large nucleus with a 
conspicuous nucleolus. Some of these cells are vacuolated. The third type has a 
round, non- vacuolated cell body 9-11 p in diameter. This cell is small as far as 
the usual neurosecretory cell is concerned. Only in the medullae terminales were 
all three types observed. The sinus gland consisted of nerve endings alone, no 
cell body having been observed. Of all the ganglia, the medullae terminales had 
the highest proportion of neurosecretory cells to ganglionic cells. 

DISCUSSION 

The eyestalks, supraesophageal ganglia, and sinus glands of Upogebia possess 
a principle that concentrates the red pigment in this animal (Fig. 1). As men- 
tioned above, Carlson (1936) and Sandeen and Baldwin (1962) working with Uca, 
and Hanstrom (1937) working with Palaemonetes as assay animals, found that 
eyestalk extracts of Upogebia were ineffective. However, eyestalk extracts of 
Upogebia were effective on Upogebia (Fig. 1). The species differences are a pos- 
sible explanation of the lack of response observed by the other investigators. The 
greater potency of extracts of supraesophageal ganglia from Upogebia compared 
with sinus gland extracts from this animal (Fig. 1) was also noted by Sandeen 
and Baldwin (1962) who assaved these extracts on Uca. 



CHROMATOPHORE CONTROL IN UPOGEBIA 29 

In Upogebia the medulla terminalis X-organ lies in the supraesophageal ganglia. 
The group of previously undescribed cells that compose the Rosette Body in the 
eyestalk of Upogebia (Figs. 2 and 3) may be homologous with the sensory pore 
X-organ described by Hanstrom (1939) in the eyestalk of many crustaceans, e.g., 
Palaemon, Crangon, and Homarus. He believed that this organ represents trans- 
formed sensory cells of a rudimentary eye papilla or sensory pore. Concentration 
of the erythrophores in Upogebia following injection of the eyestalk extract was 
presumably due to the secretory product of the Rosette Body. This secretory ma- 
terial may normally be transported by axoplasmic flow into the supraesophageal 
ganglia and from there to the sinus glands for storage. 

Nishida and Miyawaki (1954) described a holocrine gland in the eyestalk of 
two species of the anomuran Paralithodes. In Paralithodes the medulla externa, 
medulla interna, and medulla terminalis occur in the eyestalk but no sinus gland 
was observed. It may lie on the surface of the supraesophageal ganglia as in 
Upogebia. However, the structure described by Nishida and Miyawaki is not the 
Rosette Body of Upogebia. The present report constitutes the first description of 
the occurrence of an organ that presumably contains neurosecretory cells in the 
eyestalks of a crustacean whose optic ganglia do not lie in the eyestalk. Further- 
more, the arrangement of the cells is unique for a neurosecretory organ. 

As mentioned above, Miyawaki (1962) described three types of neurosecretory 
cells in the anomurans Eupagurus ochotensis and Paralithodes brevipes. The 
widths of the cell bodies were 100-130 /*, 30-60 p., and 10 p. In Upogebia three 
sizes of neurosecretory cells occur also. However, the largest type described by 
Miyawaki is much bigger than any that occurs in Upogebia. 

The problem of whether Upogebia is an anomuran or a macruran should be 
resolved. The chromatophore system of Upogebia does not allow us to decide 
between these alternatives because so little is known about chromatophore responses 
of anomurans in general that speculation would be meaningless. On the other 
hand, the neurosecretory system does offer a clue. The secondary return of the 
medullae terminales from the eyestalks to the head is typical of anomurans with 
reduced eyes rather than of macrurans (Hanstrom, 1948). 

SUMMARY AND CONCLUSIONS 

1. The eyestalks, sinus glands, and supraesophageal ganglia of the mud shrimp, 
Upogebia affinis, contain a principle that concentrates the pigment in its red 
chromatophores. 

2. A previously undescribed group of cells, with tinctorial properties character- 
istic of neurosecretory cells, occurs in the eyestalks. 

3. Because the arrangement of the cells in this structure is unique among 
neurosecretory organs, it is proposed that this structure be called the Rosette Body. 
The secretory product of these cells is probably conveyed through the optic nerve 
by axoplasmic flow to the supraesophageal ganglia and from there to the sinus 
glands. 

4. The Rosette Body may be homologous with the sensory pore X-organ of 
higher crustaceans. 



30 MILTON FINGERMAN AND CHITARU OGURO 

5. Three types of cells with tinctorial properties characteristic of neurosecretory 
cells occur in the supraesophageal ganglia. 

6. These observations were discussed in relation to the findings of other 
investigators. 

LITERATURE CITED 

BLISS, D. E., AND J. H. WELSH, 1952. The neurosecretory system of brachyuran Crustacea. 

Biol. Bull, 103: 157-169. 
BROWN, F. A., JR., AND L. M. SAIGH, 1946. The comparative distribution of two chromatophoro- 

tropic hormones (CDH and CBLH) in crustacean nervous systems. Biol. Bull., 91: 

170-180. 
BROWN, F. A., JR., AND H. H. SCUDAMORE, 1940. Differentiation of two principles from the 

crustacean sinus gland. /. Cell. Comp. Physiol., 15: 103-119. 
CARLSON, S. P., 1936. Color changes in Brachyura crustaceans, especially in Uca pugilator. 

Kimgl. fysiogr. Sallsk. Lund Forhandl., 6: 63-80. 
FINGERMAN, M., AND T. AOTO, 1962. Hormonal regulation of pigmentary effectors in 

crustaceans. Gen. and Comp. Endocrinol. Suppl., 1 : 81-93. 
GREEN, J., 1961. A Biology of Crustacea. Quadrangle Books Inc., Chicago. 
HANSTROM, B., 1937. Die Sinusdriise und der hormonal bedingte Farbwechsel der Crustaceen. 

Kungl. Svenska Vctenskap. Handl., 16: No. 3, 1-99. 

HANSTROM, B., 1939. Hormones in Invertebrates. Oxford University Press, Oxford. 
HANSTROM, B., 1948. The brain, the sense organs, and the incretory organs of the head in the 

Crustacea Malacostraca. Bull. Biol. France et Belg. Suppl., 33: 98-126. 
HOGBEN, L. T., AND D. SLOME, 1931. The pigmentary effector system. VI. The dual character 

of endocrine co-ordination in amphibian colour change. Proc. Roy. Soc. London, Ser. B, 

108: 10-53. 
MIYAWAKI, M., 1960. On the neurosecretory cells of some decapod Crustacea. Kumamoto J. 

Sci., Ser. B, 5: 1-20. 

NISHIDA, H., AND M. MIYAWAKI, 1954. Occurrence of a remarkable glandular body in the eye- 
stalk of Paralithodes camtschatica and P. brevipes. Zool. Magazine, Japan, 63 : 

274-277. 
PASSANO, L. M., 1951. The x organ-sinus gland neurosecretory system in crabs. Anat. Rec., 

Ill: 86-87. 
PERKINS, E. B., AND B. KROPP, 1932. The occurrence of the humoral chromatophore activator 

among marine crustaceans and its effect upon the chromatophores of crustaceans, fishes, 

and Amphibia. The Mount Desert Island Biol. Lab. Annual Report, pp. 24-26. 
SANDEEN, M. L, AND M. F. BALDWIN, 1962. The comparative distribution of fiddler crab 

chromatophorotropins in the burrowing shrimp, Upogebia affinis, and the isopod, Ligia 

exotica. Amer. Zool., 2: 554. 

WATERMAN, T. H., AND F. A. CHACE, JR., 1960. General crustacean biology. In: The Physiol- 
ogy of Crustacea. Vol. I, Metabolism and Growth, T. H. Waterman, ed. Academic 

Press, New York. 



THE BIOLOGY OF ASCIDIA NIGRA (SAVIGNY). 11. THE 

DEVELOPMENT AND SURVIVAL OF 

YOUNG ASCIDIANS i 

IVAN GOODBODY 
Department of Zoology, University of the West Indies, Jamaica 

In the first paper in this series (Goodbody, 1962) a description was given of 
the survival and mortality of ascidians after they first appeared as black animals 
visible to the naked eye. Because of the technique used for studying those popu- 
lations it was believed that most of them were about four weeks old when first 
recorded, and no information was available on the survival of animals of younger 
age. The work outlined in the present paper was designed to study two things : 
first, to study the development of ascidians from fertilization of the egg to the 
appearance of black pigment in the functional animal, with a view to producing a 
time scale for the different phases of development. Second, to study the survival 
of young animals from first settlement to the appearance of black pigment, and to 
determine whether there was any relationship between mortality rate and the dif- 
ferent stages of development. While this has been achieved in some measure, it 
will become apparent that a far more fundamental problem emerges from it which 
will require further study. This is the relation of changes in the biotic environment 
to patterns of survival. 

METHODS 

Populations of young ascidians were obtained either artificially in the laboratory 
or through natural settlement on glass slides in the sea. Laboratory-reared popu- 
lations were obtained through artificial fertilization of ascidian eggs : these hatched 
in 10 hours and the larvae were then pipetted gently into a perspex (lucite) settling 
box. This box, approximately 4 inches square and 2^ inches in height, has walls 
of black perspex, a base of clear transparent perspex and an open top. The box 
is designed so that four standard microscope slides can be fitted, standing up edge 
to edge along each wall, in such a way that the inside of the box is lined with glass. 
One face of each slide is gently ground with carborundum to provide a roughened 
surface for settlement ; by means of small clips at the top the slides can be secured 
with this roughened surface facing the inside of the box. With the slides in place 
and the box filled with water, the larvae were pipetted into it and the box then 
placed on a stand in an aquarium tank so that the flow of water outside the box kept 
the water inside cool. The clear base to the box allows light to come into it from 
below as well as from above : the larvae, which tend to settle on dark surfaces, 
were thus induced to settle on the glass slides backed by the dark wall of the box. 
The box was left thus for about six hours until such time as the larvae had settled. 

1 Contribution No. 313 from the Bermuda Biological Laboratory. 

31 



32 IVAN GOODBODY 

Attached larvae can be seen readily with the naked eye when a slide is withdrawn 
from the water. 

When settlement was complete the slides were transferred to an open rack in 
the aquarium and left there overnight to ensure proper attachment before they were 
further disturbed. The following day each slide was transferred in turn to a 
perspex dish mounted on the moving stage of a binocular microscope. This dish 
is designed to contain sufficient water to cover the slides adequately during inspec- 
tion, and has two projecting ledges on the inside base to enable each slide to be 
positioned in exactly the same place at successive inspections. In this way, and 
utilizing the graduated scales of the moving stage, it was possible to plot accurately 
the position of every young ascidian on each slide. When this was complete the 
slides were mounted edge to edge in a perspex frame backed by black perspex. 
(For further details of these frames, see Goodbody, 1961.) The black backing to 
the slides in the frames served two functions : it prevented any settlement of other 
organisms on the back of the slides which hence continued to fit properly into the 
microscope dish, and by providing a dark background it encouraged settlement of 
other organisms on the surface occupied by the ascidians, thus providing a natural 
community in which the ascidians could develop. These frames containing the 
slides were then suspended in the sea from floating rafts as described earlier (Good- 
body, 1962). At intervals throughout the following six weeks the frames were 
carefully removed to the laboratory in a bucket of water; the slides were removed 
to a rack in the aquarium and examined in turn to determine how many of the 
original population still survived. To avoid any possible contamination while in 
the laboratory, special containers were reserved and used only for the transport 
and handling of these frames. 

Naturally-occurring populations were also obtained on microscope slides. Clean 
slides similar to those described above were fitted into frames and placed in the 
sea, suspended from a raft. At intervals the slides were removed and inspected for 
newly settled larvae of Ascidia nigra, which can be identified at this stage (cf. 
Goodbody, 1961). Although newly settled larvae were repeatedly observed in small 
numbers, only two populations of sufficient size, and each comprising a natural 
cohort of animals derived from a single settlement, have been recorded and fol- 
lowed through their subsequent history. These two, however, provide a valuable 
supplement to the data obtained from the artificially settled populations. 

The data on the developmental stages of the young ascidian were first worked 
out at Bermuda and have been confirmed and added to from observation and 
measurement of animals within these populations. Measurements were made 
with the aid of a micrometer eye-piece. All the data on survival were obtained at 
Port Royal, Jamaica. 

THE SEQUENCE OF EARLY DEVELOPMENT 

The sequence of events leading from the fertilization of the egg to the appear- 
ance of a black ascidian, visible to the naked eye, can be divided into stages as fol- 
lows: embryonic development, the free swimming larva, settlement and meta- 
morphosis, the functional protoascidian with two protostigmata on each side, the 
six protostigmata stage, twelve stigmata, six rows of eight stigmata, twelve rows of 



BIOLOGY OF ASCIDIA NIGRA 

stigmata and developing red pigmentation, and finally the opaque black ascidian. 
This whole sequence of events takes 19 days from the fertilization of the egg, at 
temperatures in the vicinity of 27 C. Illustrations of comparable stages in young 
ascidians are to be found in the papers of Berrill (1935, 1947). 

The unfertilized egg measures about 160 ju, in diameter; fertilization is external 
and the first cleavage (under laboratory conditions) follows in about 30 minutes. 
Subsequent cleavages follow in rapid succession, and the free swimming larva 
hatches 9 to 10 hours after fertilization. Grave (1935) gives a developmental time 
of 8 hours for this species at temperatures between 27 and 29 C. The same 
author's (1925) developmental time of 6 hours and 38 minutes was recorded at the 
exceptionally high temperature of 33 C. The total length of the larva is about 
825 [*., of which 200 ^ comprise the length of the body and the remaining 625 ju. the 
tail. Larval behavior has not been studied in detail but, in common with most 
other ascidian species, the larvae of A. nigra are at first positively, and later nega- 
tively, phototropic. They are probably actively attracted towards iron, as sub- 
merged iron structures, which are not painted, often have dense growths of this 
species attached to them. Grave and Nicoll (1940) showed that iron accelerates 
metamorphosis in A. nigra. 

Settlement of larvae under laboratory conditions usually occurs within 3 to 6 
hours of hatching but has sometimes been delayed for as long as 12 hours. Grave 
and Nicoll (1940) give times ranging from 7\ to 30 hours at Tortugas, Florida, 
according to the time of year. In field conditions an extended larval life may be 
more common, particularly if the larvae take a long time to find a suitable sub- 
stratum on which to settle. When settlement is complete the tissues of the tail, 
except for the surrounding test, are resorbed into the body of the animal and 
accumulate there as a ball of nutrient reserves which nourish the growing animal 
until the gut is functional ; these reserves have usually completely disappeared 
within 24 hours of the commencement of feeding by the functional animal. In the 
first stages of metamorphosis, when the tail has just been resorbed, the animal meas- 
ures about 200 /u. in length but this increases to 250 ju, by the time the functional 
protoascidian commences feeding ; this stage is reached about 45 hours after 
settlement of the larva. 

The sequence of events from settlement of the larva is summarized in Table I. 
but it must be emphasized that there is some variation in the times and sizes given. 
A new stage of differentiation will only occur when the animal has reached a 
definite size : under conditions of poor food supply, growth may be retarded and 
hence the time scale will be correspondingly lengthened. 

The functional protoascidian has two protostigmata on each side of the 
branchial sac, a single branchial aperture, paired peribranchial apertures, and a 
functional gut with oesophagus, stomach and intestine. Simultaneously with the 
commencement of feeding the first renal vesicle can be seen, close to the oesophagus. 
During the subsequent two days there is little new differentiation but the animal 
grows from about 250 /u, to 550 p. and at the same time the nutrient reserves from 
the tail tissues completely disappears. Between the fifth and ninth days, and while 
the animal continues to grow, the original two protostigmata divide in such a way 
that there is formed, first a single row of six protostigmata, and then six rows of 
definitive stigmata on each side. There is considerable variation in the sequence 



IVAN GOODBODY 



TABLE I 

Time scale for the development of young Ascidia nigm from settlement to the appearance 

of black pigment 



Time from 
settlement 
of larva 


State of development 


Approximate 
length 

0*) 


24 hours 


Branchial siphon, 2 protostigmata, gut and heart just visible. 
Heart just commencing to beat. Large food reserves. 


200 


45-48 hours 


Functional protoascidian with 2 pairs of protostigmata, food 
in gut, 1 renal vesicle. Food reserve small. Paired peri- 
branchial apertures. 


200-300 


48-96 hours 


Growth of the protoascidian and complete disappearance of 
food reserve. No new differentiation. 


250-625 


5-6 days 


Differentiation of 6 protostigmata on each side. Branchial 
siphon develops lobes and red pigment spots between them. 
At least three renal vesicles. 


500-700 


7-8 days 


Six protostigmata divide into 6 pairs of stigmata. Peribran- 
chial apertures move toward mid-line and fuse. Four bran- 
chial tentacles visible. At least 9 renal vesicles. 


700-1150 


8-12 days 


Six rows of stigmata completed. Peribranchial siphon be- 
comes tabulated and develops red pigment spots. 


1125-1875 


13-15 days 


Twelve rows of stigmata completed. Red pigmentation 
begins to develop. 


1800-2400 


19 days 


Opaque black all over. 


2750 



of events occurring, and sometimes the protostigmata in the center of the row of 
six will have divided into three or more parts before the outer ones have completed 
their first division. On the fifth or sixth day the branchial siphon becomes lobed 
on its margins and red pigment spots develop between the adjacent lobes. The two 
peribranchial apertures begin to move towards the dorsal mid-line on about the 
seventh day, and by the eighth they have fused to form a single atrial siphon. 
Grave (1925) recorded fusion of the peribranchial apertures between the seventh 
and ninth days. 

After the ninth day there appears to be a slight pause as further growth occurs 
up to a maximum of nearly 1900 p, before the six rows of stigmata again divide to 
give twelve rows of stigmata on either side. It is at this stage, between the thir- 
teenth and fifteenth days for normally growing animals, that the first signs of 
pigmentation appear. The pigment arises as a deep red coloration all over the 
animal which gets progressively darker until the eighteenth day from settlement, 
and at a size of about 2750 ^ the animal is so black that no further details of 
internal development can be observed. 

Developmental times for other species of simple ascidian have been given by 
Berrill (1935) and Millar (1951, 1954). For AscidicUa aspersa (Miiller) and 



BIOLOGY OF ASCIDIA NIGRA 



35 



Ciona intestinalis (Linnaeus) at 16 C, Berrill records hatching after 24 hours; 
commencement of heart heat at 150 hours; appearance of functional protostigmata 
at 260 hours, and six rows of eight stigmata at 6 to 8 weeks. In comparison with 
these data for Ciona and Ascidiella, and allowing for a difference of 11 C. in the 
environmental temperature, the development of A. n'ujra seems extremely rapid. 
All the figures given are consistent with its development occurring six times faster 
than that of the other two species. Millar records the functional protostigmata 
stage as being reached in 10 days in both Pyura sqitanntlosa (Alder) and P. 
iiticrocosiints (Savigny) at temperatures in the vicinity of 20 C., which again is 
about six times as long as A. nigra takes to reach the same stage at 27 C. 

SURVIVAL 

As indicated earlier, survival has been studied in two types of poulation. In 
the first place there are data for seven different populations derived from artificial 
fertilization and subsequent settlement in the laboratory. As soon as settlement 
was complete and metamorphosis had begun, these populations were transferred to 
frames in the sea. Second, there are data from two other populations which set- 
tled naturally on microscope slides placed in frames in the sea for that purpose. 
The environmental histories of these different types of population are different, 
and comparison of the two must be made with caution. The populations derived 
from artificial fertilizations were allowed to settle on completely clean slides on 
which there were no competitors at the times of settlement ; they were, in fact, the 
first colonizers of these slides and other colonizers apeared only after the slides 
had been placed in the sea, usually about 12 hours after settlement had taken place. 
This is an unnatural situation. The other populations appeared on the slides only 
when a sessile community had already begun to develop : in one of them. Group II, 
the ascidian larvae did not settle in the community until it was 16 days old; in the 
other (I) they settled after the slides had been in the sea for only two days. 

* J J 

The data on these nine populations are presented in Tables II and III. Table 
II shows the total number of animals in each population, together with the date of 
settlement and depth at which it was reared. In Table III the survival of each 
population is presented in terms of initial cohorts of 1000 animals each. Special 
note should be taken of population H, which settled naturally but was not observed 
until the third day after settlement. At this time there were 218 animals present. 



TABLE II 

Total number of animals, date of settlement and depth at which populations were reared. 

Note that the number of animals in population H is a re-calculated figure from 218 

animals observed on third day (see text) 



Population 


A 


B 


C 


D 


E 


F 


G 


H 


I 


Total no. of animals 


80 


488 


173 


148 


72 


105 


79 


271 


34 


Date of settlement 


23.6.61 


30.6.61 


7.8.61 


7.8.61 


11.8.61 


11.8.61 


31.1.59 


24.9.61 


23.10.61 


Depth in feet 


4 


4 


4 


7 


4 


7 


7 


7 


7 


Artificial or natural 


A 


A 


A 


A 


A 


A 


A 


\ 


X 


settlement 





















36 



IVAN GOODBODY 



TABLE III 

Survival rate (Ix) in 9 populations of young Ascidia nigra, presented as cohorts of 1000 
animals each. For further data on these populations, see Table II. 



No. of 
days since 
settlement 


Population 


A 


B 


C 


D 


E 


F 


G 


n 


I 





1000 


1000 


1000 


1000 


1000 


1000 


1000 


1000 


1000 


1 


950 





960 


919 

















2 


862 


846 


873 


865 

















3 


525 


814 


815 


831 


639 


638 





803 


941 


4 


437 


773 








583 


571 


342 








5 


362 


748 

















635 





6 





619 





514 


333 


495 











7 





518 


486 





208 


448 











8 


275 











181 


400 





565 


588 


9 











365 








215 








10 


262 


434 


231 





Ill 


286 





458 





11 











223 














500 


12 


250 











55 


267 











13 





348 


145 














362 





14 


238 








162 


42 


210 











15 








87 

















353 


16 





299 





135 








76 


273 





17 








58 





42 


162 











18 


238 


279 





122 

















19 














14 


76 





177 





20 








17 




















21 


225 


240 


- 





14 


67 








176 


22 








17 


81 

















23 


225 


200 














51 








24 











61 











48 





25 





191 



















118 


26 


200 













57 











27 





















30 





28 


187 


154 


















58 


29 










61 
















30 


















38 








.31 


























32 





129 




















33 


175 












48 










.34 




















11 





35 





127 




















.36 


























37 










41 















38 


























39 





119 




















40 






















- 


41 

























42 














29 







30 


43 


175 




















44 






















45 








20 








3.7 





BIOLOGY OF ASCIDIA NIGRA 37 

The figure of 271 given in Table II is a. re-calculated value of the probable size of 
the original population when it settled. This has been calculated from the mean of 
the known survival in populations D, F and I on the third day ; these populations are 
chosen because they were reared at the same level as H, and data are available for 
the third day of life. The re-calculation introduces a possible error in the data, but 
without it we cannot compare this interesting natural population with those which 
were artificially reared. The extent of the error is not large and is confined to the 
position of the first point on the survival curve and the percentage of survivors 
remaining at the terminal point; the intervening curve would be displaced but not 
altered in form. The percentage of survivors at the end might be depressed as low 
as 0.24% or elevated to 0.45% as against 0.37% resulting from the re-calculation. 
These figures are based on the assumption that the highest mortality in the first 
three days might be equivalent to that shown by population A (47.5%), which 
would result in 0.24% of survivors at the terminal point, and the extreme possibility 
that there had been no mortality in the first three days when there would be 0.45% 
of survivors at the end ; both alternatives are improbable. 

The data are presented graphically in Figures 1 to 4. It will be apparent from 
these figures that there is not one clear-cut type, but several groups, of survival 
patterns. Populations A and B form one grouping, with 17.5% and 12% survival 
after 40 days ; populations D, F, G and I form a second grouping, with between 2% 
and 3% survival after the same period, and populations C and E form a third 
group, all of which had been lost by the twenty-fifth day. Population H, the large, 
naturally settled population, followed the pattern of D, F, G and I at first but 
diverged markedly after about 20 days ; it is discussed further below. 

These different survival patterns can. to some extent, be related to differences 
in the environment under which the populations were reared. If we consider 
populations A, B, C and E at first, we find that all four populations, derived from 
artificial fertilizations, were reared in the sea at a depth of four feet below the 
surface ; since they were suspended from a floating raft this depth remained con- 
stant, irrespective of tidal movements. Populations A and B were derived from 
fertilizations at the end of June, C and E were from fertilizations in early August. 
In early August and through September of 1961, when these experiments were 
carried out, there was a dense settlement of algal spores and subsequent growth of 
filamentous algae on all frames at the four-foot level. The frames with populations 
A, B, C and E were all equally affected by algal growth. The continued survival of 
populations A and B and the total loss of populations C and E under these condi- 
tions may be explained by the fact that the survivors in A and B were, by this time, 
quite large animals, while the survivors in C and E were still in the early stages 
of development and probably more susceptible to smothering by algae. We do not 
have sufficient data on other environmental factors at these times to be certain that 
it was the algae and not something else which caused the rapid decline of C and E. 
However, we may now compare these two populations with populations D and F, 
also derived from artificial fertilizations in early August. D and F were reared at 
7 feet below the surface, left 3% of survivors after 6 weeks and did not have the 
same dense growth of algae on the frames as did C and E. This, then, strengthens 
the conclusion that the dense algal growth may have been responsible for the 
decline in populations C and E. Populations C and D were derived from the same 



IVAN GOODBODY 



IOOO-f 



500- 



OO- 



50- 



00 

a: 

o 



ID 
GO 



IO- 



5- 



\ 



\ 



O 



1~ 
10 



1 r 



~1 r 

20 



-i P 



DAYS 



3O 



40 



FIGURE 1. Survival in two populations of young ascidians, A (- 



-) and B ( 



settled in June, 1961, and reared four feet below the sea surface. 



fertilization and were placed in the water on August 7th, the one three feet below 
the other in the water ; the deeper one, D, with little algal growth, had a higher rate 
of survival than the upper algal-covered population C. Similarly, populations E 
and F were derived from a single fertilization on August llth; F was kept three 
feet below E and had more survivors. 

The high level of survival in populations A and B in comparison with D and F 
is also of interest but not readily explained. A and B were derived from fertiliza- 
tions in late June and were maintained four feet from the surface; D and F were 
derived from fertilizations in early August and were maintained 7 feet from the 
surface. The difference in survival between the two sets of populations may be due 



BIOLOGY OF ASCIDIA NIGRA 



39 



to an environmental factor, possibly a biotic factor concerned with competition from 
other sessile organisms. Similar sorts of communities developed with each 
population dominated by cirripedes, serpulids and colonial ascidians. The data on 
the associated fauna and flora are necessarily rather superficial but there is nothing 
in the record to suggest that one community differed much from the other. Atten- 
tion should be drawn to the fact that June is one of the months when breeding in 
A. nigra is normally at a minimum (Goodbody, 1961), suggesting that this is a poor 
time of year for breeding and subsequent survival. August, on the other hand, is 
at the beginning of the autumnal rise in breeding activity of this species, suggesting 
that environmental conditions for breeding and survival of the young are improving. 



1000- 



5OO- 



00- 



5O- 



O 



a: 



10- 



5- 




O 



10 



20 



DAYS 



30 



4O 



FIGURE 2. Survival in two populations of young ascidians, C (- 



settled in August, 1961, and reared four feet below the sea surface. 



-) and E (- ), 



40 



IVAN GOODBODY 



IOOO 
500- 



IOO- 



5O- 



cr 
O 



10- 



5- 




O 



10 



20 



DAYS 



3O 



4O 



FIGURE 3. Survival in three populations of young ascidians reared 7 feet below the sea 

surface. D (- -) and F (- ) were settled in August, 1961. G ( ) 

was settled in January, 1959. 

It is surprising, therefore, to find that the June population shows such a high level 
of survival. 

There remain three other populations to be discussed. G was a population set 
up in January, 1959, and maintained 7 feet below the surface. It was followed 
for only 30 days and until then showed a survival pattern closely similar to those 
populations settled in August and maintained at 7 feet. 

Finally, there are data on two populations derived from fertilizations in the 
sea, which settled naturally on slides at 7 feet. H was a large population settling 
on slides which already had two weeks' growth of barnacles and other sessile 
forms. I settled two days after slides had been placed in the water and is therefore 



BIOLOGY OF ASCIDIA NIGRA 



41 



more nearly comparable to the artificial settlements ; it is, however, a small popula- 
tion compared to any of the others. 

Population I shows a survival curve closely approximating that of the artificial 
populations D, F and G, reared at 7 feet, and like them had 3% survival at the 
end of 6 weeks. However, it must be emphasized that this represents no more 
than one animal surviving. Population H follows the pattern of D, F, G and I 
until about the twentieth day, when the curve falls sharply, and the final survival at 
the forty-fifth day is only 0.37%, but this also represents only one surviving animal. 
Nevertheless, the curve merits some attention, as there is a possible explanation for 
the divergence between it and the others. The animals in this population settled in a 




3O 



4O 



DAYS 



FIGURE 4. Survival in two naturally settled populations of young ascidians, 7 feet below the 

sea surface. H ( ) settled in September, 1961, and I ( ) settled in October, 

1961. The data for H have been re-calculated to obtain the first point on the curve (see text). 



42 IVAN GOODBODY 

population which was already 16 days old, whereas those in population I settled in 
one which was two days old, and D, F and G were first colonizers. We might 
expect therefore that competition from organisms already in possession of the 
slides would cause heavy mortality in population H, but we would also expect 
this to occur in the youngest stages from about the second day to the nineteenth day. 
that is, between the onset of feeding and the appearance of a fully formed black 
ascidian. This, in fact, does not appear to have happened and during this phase of 
development the survival curve follows closely that of the populations which 
settled early in the life of the community. The divergence between the curves does 
not take place until the point at which the animals had reached the opaque black 
stage. We must, therefore, look for another explanation, and this may be found in 
the nature of the community which developed in association with this population. 
The usual community developing in association with these populations comprised 
green algae (Enter omorp ha) , cirripedes (Balanus), serpulid polychaetes (Eupo- 
mahis), colonial ascidians (Dideinnum candidum Savigny), Diplosoina inac- 
donaldi Herdman and Symplegma viride (Herdman), lamellibranchs (probably 
Ostrea equcstris Say) and a large errant fauna of amphipods, copepods, turbel- 
larians, polychaete annelids, herbivorous gastropods (on the algae) and occasionally 
nudibranchs. The community in association with population I differed markedly 
from all others in that a large population of pyurid ascidians (Microcosnius exas- 
peratus Heller) developed and was already in evidence by the time the Ascidia 
nigra were 20 days old ; they eventually became the dominant organism in this 
community. Pyurid ascidians in general belong to a later stage in the natural 
succession of inshore sessile communities in Jamaica (unpublished observations), 
and may in this case have played some part in displacing the A. nigra, either by 
competition for food or by squeezing them off their attachment bases. 

CONCLUSIONS 

The general picture emerging from these data on Ascidia nigra is that survival 
of between 2% and 3% may be normal for a population at 7 feet after the first 6 
weeks of life. All the artificial populations and one of the natural populations reared 
at 7 feet below the sea surface exhibited this pattern of survival, but this is not 
true of populations reared at the four-foot level. Of four populations reared at this 
level, two left no survivors and two left between 12% and 18% survivors. 

It is probable that the two populations which left no survivors were affected 
adversely by dense algal growth, but until more information is available, we can 
only speculate on the high survival of the other two. This extreme divergence 
between growth of populations reared at the four-foot level, and the smaller 
divergence between population H and the other populations reared at 7 feet, does, 
however, focus attention on the effects of environmental differences on survival rates 
of developing ascidians. These differences can be considered as having dimensions 
in both space and time. On the one hand, divergence between populations at the 
four-foot level is assumed to be the result of seasonal differences in the settlement 
of other organisms at that level. On the other hand, the difference noted between 
populations at different vertical levels is probably indirectly due to differences in 
illumination ; this results in different biotic communities developing in association 
with, and in competition with, the ascidians. In both cases it is assumed that dif- 



BIOLOGY OF ASCIDIA NIGRA 43 

ferences in the associated community are responsible for the differences in survival, 
but the community differences are due in the first instance to seasonality and in 
the second to spatial separation. Similarly, it is suggested that the divergence 
between the two natural populations, H and I, is a biotic effect due to direct compe- 
tition between two species of ascidians. These are obviously very complex prob- 
lems and we hope to progress further with them in the next few years. 

At the outset of this work it was thought possible that there might be some 
correlation between mortality rate and the stage of development of the young 
ascidian. There is no clear-cut evidence of such a correlation from these data, 
and the most that can be said is that the highest mortality occurs before the 
animal becomes completely black, about the nineteenth day. The causes of mortality 
in the young ascidian are varied. Some zooids are undoubtedly eaten by flatworms 
and occasionally by young polychaete worms ; others have been found dead on the 
slide with numerous ciliate protozoa inside the empty test. Other ascidians appear 
to starve to death in the shadow of a barnacle ; ascidians growing very close to 
barnacles have been observed to shrivel slowly from one inspection to another, as if 
they were unable to obtain sufficient food in competition with the barnacle. A 
further cause of mortality is due to spatial competition with colonial ascidians such 
as Diplosoma macdonaldi, Didcmmnn candidmn and Symplcgnia viride ; all these 
species have been seen to smother young A. nigra as the colony spreads across the 
slide. Mention has already been made of the possibility that dense mats of algal 
growth may be responsible for ascidian mortality, presumably by preventing an 
adequate flow of food from reaching the ascidian. 



This work was originally supported by a grant from the Nuffield Foundation, 
to whom grateful acknowledgment is made. Much of the original observation on 
the early developmental stages was made at the Bermuda Biological Laboratory ; 
I am grateful to the Director and staff of that laboratory for their help there, and 
to the National Science Foundation and the University College of the West Indies 
for grants making the visit possible. I am also grateful to Mr. B. Wade who 
checked many of the populations during the summer of 1961. 

SUMMARY 

1. A time scale for the development of Ascidia nigra is given, from fertiliza- 
tion of the egg to the completion of black pigment formation. Embryonic and 
larval development are completed in less than 24 hours, and a functional ascidian 
is developed within 48 hours of larval settlement. Pigment first appears about the 
thirteenth day and the animal is completely black by the eighteenth or nineteenth 
day. The whole developmental process is completed about six times faster than in 
temperate- water ascidians at temperatures about 11 C. less. 

2. The survival of populations of ascidians has been followed from settlement 
to the end of the sixth week of life. Most populations left only 2% to 3% sur- 
vivors, but some left none and one as many as 17.5% survivors. These differences 
are discussed and are assumed to be due to differences in the associated biotic 
community arising from differences in illumination and season of growth. 



44 IVAN GOODBODY 

3. Death is due to a number of factors, including predation by flatworms and 
polychaete annelids, competition for food and for space with other sessile animals 
or algae, and possibly in some cases protozoan infection. 

LITERATURE CITED 

BERRILL, N. J., 1935. Studies in Tunicate development. III. Differential retardation and 

acceleration. Phil. Trans. Roy. Soc. London, Ser. B, 225: 255-336. 

BERRILL, N. J., 1947. The development and growth of Ciona. J. Mar. Biol. Assoc., 26: 616-625. 
GOODBODY, I., 1961. Continuous breeding in three species of tropical ascidian. Proc. Zool. Soc. 

London, 136: 403-409. 
GOODBODY, I., 1962. The biology of Ascidia nigra (Savigny). I. Survival and mortality in an 

adult population. Biol. Bull, 122: 40-51. 
GRAVE, C, 1925. Preliminary report on development and behavior of larval ascidians and 

periodicity in spawning of certain marine invertebrates. Carncf/ic Inst. Wash. Year 

Book, 24: 224-228. 
GRAVE, C., 1935. Metamorphosis of ascidian larvae. Publ. Carnegie Inst. IV ash. No. 452: 

209-292. 
GRAVE, C., AND P. A. NICOLL, 1940. Studies on larval life and metamorphosis in Ascidia nigra 

and species of Polyandrocarpa. Papers Tortugas Lab. Carnegie Inst., 32: 1-46. 
MILLAR, R. H., 1951. The development and early stages of the ascidian Pyura squamiilosa 

(Alder). /. Afar. Biol. Assoc., 30: 27-31. 
MILLAR, R. H., 1954. The development of the ascidian P \itra microcosmns (Savigny). /. Mar. 

Biol. Assoc., 33: 403-407. 






CHLORIDE EXCHANGES IN RAINBOW TROUT (SALMO 

GAIRDNERI) ADAPTED TO DIFFERENT 

SALINITIES * 

MALCOLM S. GORDON 2 
Department of Zoology, Unk'crsity of California, Los Angeles 24, California 

The water and salt balance mechanisms used by teleost fishes have attracted 
considerable interest for almost a century (recent general reviews are: Black, 1957; 
Gordon, 1963; Prosser and Brown, 1961; Shaw, 1960). While a great deal of 
work has been done on fishes maintained under fairly constant conditions in either 
fresh or sea water, very little effort has been directed at studies of the changes in 
the basic regulatory mechanisms which occur when fishes encounter environ- 
mental changes. Salinity changes are among the commonest of these variations. 

The responses of euryhaline teleosts are particularly interesting in this regard, 
since these forms are usually very good osmotic and ionic regulators. The relative 
constancy of the concentration of their internal medium means that, when these 
fishes enter environments of different salinities, they maintain across their integu- 
ments osmotic gradients of different magnitudes and even directions. These 
changes in osmotic gradients in turn mean that, in order for constancy of internal 
concentration to continue, the fishes must change either or both the fluxes and ef- 
fective permeabilities for water and at least sodium and chloride across their major 
exchange pathways the gills, gut and kidneys. 

The three-spined stickleback (Gastcrostciis aculcatus] seems to be the only 
euryhaline fish for which even limited data are available on the mechanism of 
adaptation to changes in trans-integumentary osmotic gradients. Heuts (1942) 
found that the chloride concentration in cloacal excreta increased about 20 times 
over fresh-water levels in sticklebacks in l / 3 sea water. Mullins (1950) presented 
data which he interpreted as indicating that sticklebacks increased their rate of 
drinking of external medium as the concentration of that medium increased. He 
also thought that the permeability of the gills to specific ions, especially potassium, 
changed. 

The rainbow trout (Saliuo gairdneri} is a euryhaline teleost which survives 
well in both fresh water and the sea. It is a very good osmotic and ionic regulator 
(Busnel, 1942; Busnel and Drilhon, 1944. 1946; Houston. 1959, 1961; Parry, 
1960, 1961). This paper describes a study of the rates of exchange of body chloride 
by rainbow trout acclimatized to fresh water and sea water, and to several inter- 
mediate salinities. The results are interpreted as indicating that the change-over 
from hyper-osmotic regulation in fresh water to hypo-osmotic regulation in sea 

1 These studies have been supported by grants from the National Science Foundation 
(G-8802; G-238S5). 

2 Valuable technical assistance was provided by Cynthia Rosenblum, Gordon Engel, Maria 
Polak and Melvin Soloff. 

45 



46 MALCOLM S. GORDON 

water, at least in this species, is based in large part on changes in permeability to 
water of the gills and other permeable portions of the integument. 

A preliminary discussion of some of these results was given by Gordon (1959a). 

MATERIALS AND METHODS 

Sexually mature two- to three-year-old rainbow trout, of 100-250 gm. weight, 
were obtained from a commercial fish hatchery. These fish were maintained in 
individual aquaria in the laboratory. Each aquarium was equipped with aeration 
and a system whereby circulating water of any salinity between fresh water (de- 
chlorinated Los Angeles tap water) and 100% sea water (equals 32% salinity, 
960 mOsmoles/liter osmotic concentration) could be supplied continuously for any 
desired period. Water temperature was maintained at 19 2 C. Photoperiod 
was that normal for the season. No experiments were done during the autumn 
breeding season, when salinity tolerance in this species is sometimes markedly 
decreased. Fish were not fed. 

Groups of fish were maintained in fresh water for several days after receipt. 
Ten days were allowed for acclimatization to each higher salinity used. Transfers 
were made directly from fresh water to y~ and % sea water. Acclimatization to 
still higher salinities was done stepwise {e.g. 10 days in % sea water, then 10 days 
in % sea water, then 10 days in % sea water). 

Estimates of rates of chloride exchange were carried out on groups of five 
trout acclimatized to fresh water, %, V z , %, and % sea water. Excepting only 
Yf sea water, observations were made at each concentration, both on normal, intact 
fish and on fish which had had their cloacas and urinary papillae tightly ligated. 
The effectiveness of cloacal ligation was tested at the end of each experiment by 
pressing on the abdomen of the fish in order to extrude bladder urine or gut 
contents. In all cases pressures required were so far above those which the fish 
themselves could have produced that I am confident no leakage occurred during 
the experiments. All operations on trout were performed with the fish held under 
water in a piece of smooth cotton cloth. Water used was always that to which the 
fish were acclimatized. 

Radioactive chlorine-36 was used as a tracer for chloride movements. This 
isotope was obtained as carrier-free HC1 36 with specific activity of about 500 /tc./gm. 
chloride, from Oak Ridge. This solution was neutralized with NaOH and diluted 
to produce a final injection solution of about 160 mM NaCl 36 . Final activities of 
injection solutions were in the range 20-30 /xc./ml. 

NaCl 36 solutions were injected intraperitoneally. Total injection volumes were 
in the range 0.08-0.15 ml., calculated to produce doses of approximately 15 /AC. /kg. 
The fish were then transferred to individual small closed plastic aquaria, each con- 
taining 1500 ml. of water of the appropriate concentration and supplied with 
aeration. Two-mi, aliquots of the external medium were taken at one-hour inter- 
vals for 6 hours. The fish were then sacrificed, a blood sample was taken by 
heart puncture, and they were re-weighed to the nearest gram. 

The aliquots of external medium, also duplicate 25-jul. aliquots of blood samples, 
were evaporated to dryness on aluminum planchets. Times required for the oc- 
currence of 5120 counts per sample were determined with a Nuclear-Chicago Model 
D-47 thin end window (0.1 mg./cm. 2 ) gas flow counter connected with an automatic 



CHLORIDE EXCHANGES IN RAINBOW TROUT 47 

sample changer, sealer and printing timer. Background was stable at 15 cpm., 
resulting in assay precision of 2-3%. Absolute activity of samples was de- 
termined by comparison with standards of known activity made from the original 
HC1 36 solutions obtained from Oak Ridge. These standards were dissolved in ap- 
propriate volumes of the sea water dilutions used, and prepared and counted in the 
same way as unknowns. 

Total chloride in blood samples was determined on duplicate 0.1-ml. aliquots 
with an Aminco-Cotlove automatic chloride titrator. Precision was 2-3 meq./l. 
of whole, hemolyzed blood. Blood specific activity for each fish at the end of each 
experiment was calculated from these data and the radioactivity assays. 

Measured rates of appearance of Cl 36 were converted to p.c. Cl 36 /kg. fish/hr. for 
fish with a uniform specific activity of their blood one hour after injection of 0.50 
juc. Cl 36 /meq. total Cl in blood. These adjustments were required since there were 
variations in radioactivity of body chloride in each fish in each experimental group, 
which resulted from errors in initial weighing and variations in amount of total 
body chloride. 

Procedure for adjusting rate measurements was as follows: No recycling of 
radioactivity into the fish was assumed, as was a steady-state for the chloride content 
of the fish. The fraction of the injected dose which appeared in the medium during 
the 6 hours duration of the experiment was calculated. Those few fish were dis- 
carded which had exchanged more than about 20% of the injected dose (these 
probably had been damaged in handling), as were those with such low blood 
chloride specific activities that it was apparent the initial injection had not been 
successful. Rates of exchange of radioactivity were assumed to have been propor- 
tional to blood chloride specific activity. Blood chloride specific activity one hour 
after the initial injection was assumed to have been equal to : measured final specific 
activity /fraction of injected isotope dose remaining in fish at end of experiment. 
A factor was then calculated for each fish which would adjust the calculated blood 
specific activity one hour after the injection to 0.50 /xc./meq. Cl. The measured 
rates of appearance of radioactivity for each fish, in /xc./kg. /hr., were then 
multiplied by this factor. 

Data for the first hour following the injections were not used. Mean rates of 
appearance, also standard errors, for each of the remaining five hours of each 
experiment were calculated. Excepting only the experiments in fresh water, there 
were no statistically significant differences within the sets of five hourly means for 
each experiment. 

Cumulative rates of isotope appearance for the five usable hours of each experi- 
ment were next calculated for each fish in each group. Means and standard errors 
for these cumulative rates were calculated and comparisons between the various 
groups made by analysis of variance. 

As a check on the osmotic and ionic regulatory abilities of the rainbow trout 
used in this work, blood samples were taken by heart puncture from additional 
groups of animals, acclimatized to various salinities under the same conditions as 
the experimental animals. No samples were taken during the breeding season. 
Freezing point depressions were determined by the method of Ramsay and Brown 
(1955), with a precision of 0.02 C. Chloride concentrations were determined 
on the automatic chloride titrator as described above. 



48 



MALCOLM S. GORDON 



RESULTS 

Plasma osmotic and chloride concentrations in rainbow trout acclimatized to 
various salinities from fresh water to % sea water are shown in Figure 1. No 
seasonal changes were noted. These results agree with those of the other workers 
cited in the introduction. Plasma chloride concentration closely parallels osmotic 
concentration. Regulation of these concentrations is not perfect, but it is quite 
good. Fish in fresh water maintain osmotic gradients across their integuments 
almost twice as large as those maintained by fish in Yi sea water. These gradients 
act to move water into the animal. Fish in % sea water maintain osmotic gradients, 
in the opposite direction, 30 to 50 times larger than those maintained by fish in 
Ys sea water. 

Table I and Figure 2 present the results of the adjusted measurements of 
cumulative rates of Cl 36 appearance from variously acclimatized trout, both un- 
operated fish and those with ligated cloacas and urinary papillae ("ligated fish"), 
during the period February through May. Table II summarizes the results of the 
statistical analyses of these data. The measurements on ligated fish are interpreted 
as estimates of the rates of chloride exchange across only the gills and other 
permeable parts of the integument. The differences between these rates and those 
for unoperated animals are considered estimates of the rates of chloride loss by way 
of the gut and, especially, the kidneys. 



\ 



700 r 



600 



500 

Ct: 400 

PN* 

^ 300 

o 

^O 200 



100 

5 

- o 




(4) 



(4) 



200 f 400 600f 800 

EXTERNAL CONCENTRATION (mOsm/l) 



/MOOO 



FIGURE 1. Osmotic and chloride concentrations of plasma vs. osmotic concentration of 
external medium, in rainbow trout acclimatized to various salinities. Horizontal lines indicate 
means of observations on indicated numbers of fish ; vertical lines 2 S.E.'s. Solid line : plasma 
osmotic concentration (freezing point depression); dashed line: plasma chloride concentration. 
Diagonal line is line of equality. Fish acclimatized to each medium for at least 10 days. All 
experiments at 19 2 C. Arrows along abscissa indicate osmotic concentrations of fresh water, 
%, % and % sea water. 



CHLORIDE EXCHANGES IN RAINBOW TROUT 



12 r 




| 200 | 400 600| 800 ^ 1000 

EXTERNAL CONCENTRATION (mOsm/l) 

FIGURE 2. Cumulative rate of appearance of Cl 36 from rainbow trout acclimatized to 
various salinities at 19 2 C. between February and May. See text for details. Solid line : 
unoperated fish; dashed line: fish with sewn cloacas and urinary papillae. Symbols as Figure 1. 



Rates of total chloride exchange for unoperated animals in hypo-osmotic media 
(fresh water and % sea water) are all statistically significantly lower than rates for 
unoperated animals in all of the hyper-osmotic media. Total exchange rates 
increased as external osmotic and chloride concentrations rose. Rates for fish in % 
sea water are not statistically significantly different from rates for fish in % sea 
water. The difference in rates between % and % sea water is significant. 

Fish in fresh water are indicated to have exchanged chloride about one-half 
as rapidly as fish in y 7 sea water. The external chloride pool in the fresh-water 
experiments was sufficiently small so that this result may have been affected by- 
recycling of isotope by the fish. There was a statistically significant continuous 
secular decline in rate of change of external isotope concentration in the fresh-water 
experiments. The rate of isotope appearance during the second hour of the fresh- 
water series was the same as the rate determined for the same period in % sea water. 
The rate constants calculated for these two groups by analyses of the data as two 
compartment systems (Solomon. 1960, p. 130) were also the same. The true 



50 



MALCOLM S. GORDON 



TABLE I 

Cumulative rates of Cl 36 appearance from rainbow trout (blood specific activity 
0.50 tic. Cl^/meq. Cl at one hour; 19 2 C.) 



State of 
acclimatization 

Unoperated fish 

FW 

1/7 SW 
1/3 SW 
2/3 SW 
3/3 SW 

Fish with ligated cloaca 
and urinary papilla 

FW 

1/3 SW 
2/3 SW 
3/3 SW 



Rates of Cl 36 appearance (pc./kg./5 hr.) 
[A' S.E. (N)] 



February-May 

0.51 0.06 (5) 
0.98 0.24 (4) 
2.16 0.42 (4) 
5.16 1.09 (4) 
9.6 1.2 (2) 



0.61 0.22 (3) 

1.75 0.48 (3) 

0.54 0.20 (4) 

2.72 0.24 (3) 



July-August 



1.75 0.16 (4) 
3.68 0.59 (4) 
2.24 0.17 (3) 



cumulative rate for fresh water is therefore probably similar to the rate for % 
sea water. 

Fish in % sea water exchanged chloride approximately 4i times more rapidly 
than fish in % sea water, about twice as rapidly as fish in % sea water. 

The pattern for ligated fish is statistically significantly different from that for 
unoperated fish only in % and % sea water. The results for the ligated group in 
fresh water duplicated those for unoperated fish both in time sequence and cumula- 
tively. This indicates that virtually all chloride exchange in trout in fresh water 
occurs across the gills and other permeable parts of the integument. The same 
appears to be true for fish in % sea water. 

The situation in strongly hyperosmotic media appears to be different. Rate of 
chloride exchange by ligated fish in % sea water was only about 10% of that for 
unoperated fish. In % sea water this fraction was near 25%. The rate for ligated 
fish in % sea water is not statistically significantly lower than the rate for ligated 
fish in y s sea water. The rates for ligated fish in % sea water and % sea water 
differ significantly. The rates for ligated fish in % sea water and % sea water are 



TABLE II 

' F" values resulting from analysis of variance comparisons between groups of 
rainbow trout, February through May* 



Unoperated fish 


FW 


1 /7 SW 


1 /3 SW 


2/3 SW 


3/3 SW 


Ligated fish 


3/3 SW 


396*** 


297*** 


133*** 


13.0*** 





3/3 SW 


2/3 SW 


12.1*** 


9.59** 


4.54 





61.7*** 


2/3 SW 


1/3 SW 


12.4*** 


5.35** 





2.50 


1.40 


1/3 SW 


1 /7 SW 


3.95 














1/7 SW 


FW 


~ 





2.20 


0.30 


55.6*** 


FW 



"F" values with no asterisks indicate that the groups compared are not statistically sig- 
nificantly different. Two asterisks indicate significance at the 5% level, three asterisks at the 
1% level. 



CHLORIDE EXCHANGES IN RAINBOW TROUT 



51 



15 r 



\ 



10 



Uj 



^ F 

II* 



(2) 




(4) 



4 

(3) 



200 



400 



600f 



800 



1000 



EXTERNAL CONCENTRATION (mOsm/l) 



FIGURE 3. As Figure 2, but unoperated fish at different times of year. Solid line : 
experiments between February and May ; dashed line : experiments in July and August. 

not statistically significantly different. These results indicate that trout in hyper- 
osmotic media may exchange by way of the permeable parts of their integument 
only a small fraction of the total amount of chloride they take in by drinking of 
external medium. The amount of chloride exchanged across these permeable areas 
(presumably primarily the gills) does not change greatly, even though the osmotic 
gradients maintained by the fish increase tremendously. 

Table I and Figure 3 present the results of the adjusted measurements of 
cumulative rates of Cl 36 appearance from variously acclimatized unoperated trout 
during the periods February through May and July through August. A compari- 
son of these two periods was made because Gordon (1959b) had found the osmo- 
regulatory abilities of the brown trout (Salmo trutta} to be significantly lessened 
during July and August, as compared with the rest of the year. 

The July-August results differ from the February-May results only for fish in % 
sea water. Trout acclimatized to % sea water in summer exchanged chloride only 
% as rapidly as similarly treated fish in spring. The lack of similar differences for 
trout in y 3 and % sea water makes it improbable that this single difference actually 
reflects seasonal changes in the fish. It seems much more probable that the dif- 
ference was due to differences in experimental manipulation of the two groups, the 
summer group having been less shocked and damaged by handling than the spring 
group. 

DISCUSSION 

Two major difficulties complicate the interpretation of the results of experi- 
ments such as these. First, fishes are notoriously sensitive to handling. This is due 
not only to the changes in internal hormonal concentrations resulting from the 



52 MALCOLM S. GORDON 

stress imposed by experimental manipulations, but also to mechanical changes in 
the permeability of their integument due to rubbing off of mucus, scratches, etc. 
The usual result of these effects is the induction, apparently only in fish in hyper- 
osmotic media, of a more or less severe state of "laboratory diuresis." This condi- 
tion is characterized by the production of abnormally large volumes of urine of 
abnormally high salt content. The condition appears almost immediately after 
handling of fish (Forster and Berglund, 1956; Holmes, 1961). 

It is probable that the increased rates of chloride exchange shown by unoperated 
rainbow trout in % and % sea water were due in large part to the occurrence of 
progressively more severe cases of laboratory diuresis. The considerable variability 
of most of the measurements for such groups fits in with this interpretation. The 
low result for the group in % sea water in summer is probably the most reliable of 
the four presented. It seems most reasonable, therefore, to consider all of these 
results as upper limits for the total rates of chloride exchange by the fish involved. 

The second complication is that an unknown and perhaps variable fraction of 
the measured rates of isotope appearance may have been due to the physical ex- 
change of unlabelled for labelled Cl atoms ("exchange diffusion") and not have 
represented actual unidirectional ionic fluxes produced by particular transport 
processes. None of the data required for estimation of the rates at which exchange 
diffusion may have taken place are available (Cooperstein and Hogben, 1959). It 
is therefore impossible to say with certainty that the rates measured for ligated fish 
represent either (a) rates of passive outward diffusion of chloride across the gills, 
etc., of trout in hypo-osmotic media, or (b) rates of active excretion of chloride 
across the same tissues of fish in hyper-osmotic media. The rates determined are, 
however, upper limits for the rates at which these processes could occur. 

Even with these qualifications, several inferences still seem reasonable. Note 
again that in both hypo-osmotic and hyper-osmotic media neither total chloride ex- 
changes nor integumentary chloride exchanges varied proportionally with the 
magnitudes of the trans-integumentary osmotic gradients maintained by the fish. 

Assume that the surf ace/ volume ratio for the trout is constant in all salinities. 
Let the null hypothesis be that the diffusion coefficients for water across the 
permeable parts of the integument of rainbow trout are constant in all media. In- 
creases in magnitude of osmotic gradients should, in this situation, produce pro- 
portional increases in rates of water movement. Assuming the experimental fish 
were in steady states with regard to water and salt, the implications would be: (a) 
the more dilute the external hypo-osmotic medium, the greater the rate of urine pro- 
duction and, with fairly constant urinary salt content such as usually occurs in fishes, 
the more rapid the rate of urinary salt loss; (b) the more concentrated the ex- 
ternal hyper-osmotic medium, the greater the rate of drinking of that medium 
and the more rapid the rate of active salt excretion by the gills (osmotic gradients 
maintained by trout in % sea water were a minimum of 30 times larger than 
gradients maintained by fish in y s sea water ; there is three times as much salt 
per unit volume in % sea water as there is in % sea water; trout in % sea water 
might, therefore, be expected to excrete chloride 90 or more times faster than fish 
in % sea water). 

The data agree with neither of these predictions. It seems most improbable that 
the amount of integumentary exchange diffusion in hyper-osmotic media would 



CHLORIDE EXCHANGES IN RAINBOW TROUT 53 

change sufficiently to mask changes in rates of active excretion of the magnitudes 
required by the model. It is probable, therefore, that a most important part of the 
process of salinity acclimatization in rainbow trout is a reduction in integumentary 
permeability to water. These permeability reductions are more or less proportional 
to the magnitudes of transintegumentary osmotic gradients, whatever the direction 
of these gradients. 

This conclusion implies that adaptation to different salinities does not neces- 
sarily impose on rainbow trout markedly different energy requirements for osmo- 
regulatory purposes. It is possible that much of the adjustment is taken care of 
by changes in the physical state of the permeable areas of the integument, changes 
mediated, perhaps, by a neurohypophysial hormone (Hays and Leaf, 1962). 

Support for these results and inferences can be derived from some calculations. 
Assume that the cumulative rate of chloride exchange by trout in fresh water was 
actually that measured for fish in % sea water. Assume also that the specific ac- 
tivity of the chloride exchanged was the same as the adjusted specific activity used 
for the blood, i.e., 0.50 p.c. Q 36 /rneq. total Cl. The total rate of chloride loss from 
trout in fresh water, on this basis, was about 9.6 meq. Cl/kg./24 hr. Krogh ( 1937) 
estimated that small rainbow trout in fresh water absorbed Cl from their environ- 
ment at the rate of 7.2 meq. Cl/kg./24 hr. Phillips et al. (1958), working with 
small brook trout (Salvelinus fontinalis), measured a total rate of Cl uptake from 
fresh water of 3 mM Cl concentration of about 1.8 meq. Cl/kg./24 hr. 

The data of Krogh (1937) and Holmes (1961) indicate that urinary Cl losses 
in fresh water should account for almost half of the total losses. The present data 
do not indicate this. 

Other euryhaline species, such as the common eels (Anguilla spp.) drink 30-200 
ml./kg./24 hr. when acclimatized to sea water (Smith, 1930; Keys, 1933). As- 
suming no exchange diffusion and a specific activity of body Cl of 0.50 /AC. /meq., the 
maximum measured rate of total Cl exchange in % sea water (February-May) is 
equivalent to the ingestion by the trout of about 200 ml./kg./24 hr. The rate for 
July- August fish is equivalent to about 40 ml./kg./24 hr. 

Similar calculations for trout in % and % sea water give drinking rates of 
approximately 40 ml./kg./24 hr. and 100 ml./kg./24 hr., respectively, between 
February and May. Summer fish are similar. 

SUMMARY 

1. Estimates of rates of exchange of body chloride, both total exchanges and ex- 
changes across the integument and by way of the gut and kidneys, have been made 
in rainbow trout (Salmo gairdneri) acclimatized to various salinities between 
fresh water and sea water (salinity 32% ). Radioactive chlorine-36 was used as a 
tracer of chloride movements. 

2. Neither total Cl exchanges nor integumentary exchanges varied in proportion 
with changes in the magnitude of the transintegumentary osmotic gradients main- 
tained by the fish. This result is interpreted as indicating that changes in the 
permeability to water of the integument (probably primarily the gills) are an 
important part of the salinity adaptation process in rainbow trout. 

3. Laboratory diuresis and exchange diffusion of chloride are discussed as 
possible complications in this interpretation. 



54 MALCOLM S. GORDON 

LITERATURE CITED 

BLACK, V. S., 1957. Excretion and osmoregulation. In: The Physiology of Fishes ( M. E. 

Brown, ed.), 1: 163-205. 
BUSNEL, R. G., 1942. Recherches de physiologic appliquees a la pisciculture : a propos de la 

migration de la truite arc-en-ciel. Bull. Francois Peche Piscicnlt., 15: 45-65. 
BUSNEL, R. G., AND A. DRILHON, 1944. Modifications de la concentration moleculaire du milieu 

interieur de Salmo iridaeits au cours de la croissance, en rapport avec 1'adaptation aux 

changements de salinite. C. R. Soc. Biol, 138: 334-336. 
BUSNEL, R. G., AND A. DRILHOX, 1946. Recherches sur la physiologic des Salmonides. Bull. 

Inst. Oceanogr. (Monaco), No. 893: 1-7 
COOPERSTEIN, I. L., AND C. A. M. HoGBEN, 1959. Ionic transfer across the isolated frog large 

intestine. /. Gen. Physiol, 42: 461^473. 

FORSTER, R. P., AND F. BERGLUND, 1956. Osmotic diuresis and its effect on total electrolyte dis- 
tribution in plasma and urine of the aglomerular teleost, Lophius americamts. J. Gen. 

Physio!,, 39: 349-360. 
GORDON, M. S., 1959a. Rates of exchange of chloride ions in rainbow trout (Salmo gairdneri) 

acclimated to various salinities. Anat. Rec., 134: 571-572. 
GORDON, M. S., 1959b. Ionic regulation in the brown trout (Salmo tnttta L.). /. Exp. Biol., 

36: 227-252. 
GORDON, M. S., 1963. Water and salt balance in fishes and amphibians. Handbook of Physiology, 

4 (in press). 
HAYS, R. M., AND A. LEAF, 1962. The state of water in the isolated toad bladder in the 

presence and absence of vasopressin. /. Gen. Physiol., 45: 933-948. 
HEUTS, M. J., 1942. Chloride-excretie bij Gasfcrostctis acnlcatus L. Ann. Soc. Roy. Zool. Bclg., 

73: 69-72. 
HOLMES, R., 1961. Kidney function in migrating salmonids. Ann. Kept. Challenger Soc., 3, 

No. 13 : 23. 

HOUSTON, A. H., 1959. Osmoregulatory adaptation of steelhead trout (Salmo gairdneri Rich- 
ardson) to sea water. Canad. J. Zool., 37: 729-748. 
HOUSTON, A. H., 1961. Influence of size upon the adaptation of steelhead trout (Salmo 

gairdneri) and chum salmon (Oncorhynchus keta) to sea water. J. Fish. Res. Bd. 

Canada, 18: 401-415. 
KEYS, A., 1933. The mechanism of adaptation to varying salinity in the common eel and the 

general problem of osmotic regulation. Proc. Roy. Soc. London, Ser. B, 112: 184-199. 
KROGH, A., 1937. Osmotic regulation in fresh water fishes by active absorption of chloride ions. 

Zeitschr. vergl. Physiol., 24: 656-666. 
MULLINS, L. J., 1950. Osmotic regulation in fish as studied with radioisotopes. Acta Physiol. 

Scand,, 21: 303-314. 
PARRY, G., 1960. The development of salinity tolerance in the salmon, Salmo salar (L.) and 

some related species. /. Exp. Biol., 37: 425-434. 
PARRY, G., 1961. Osmotic and ionic changes in blood and muscle of migrating salmonids. 

/. Exp. Biol., 38: 411-427. 
PHILLIPS, A. M., JR., H. A. PODOLIAK, R. F. DUMAS AND R. W. THOESEN, 1958. Absorption 

of dissolved chloride by brook trout. Cortland (N.Y.) Fish Hatchery Fish Res. Bull., 

No. 22 : 8-16. 
PROSSER, C. L., AND F. A. BROWN, JR., 1961. Comparative Animal Physiology. W. B. Saunders 

Co., Philadelphia ; second edition. 

RAMSAY, J. A., AND R. H. J. BROWN, 1955. Simplified apparatus and procedure for freezing- 
point determinations upon small volumes of fluid. /. Sci. Instr., 32: 372-375. 
SHAW, J., 1960. The mechanisms of osmoregulation. In: Comparative Biochemistry ( M. 

Florkin and H. S. Mason, eds.), 2: 471-518. 
SMITH, H. W., 1930. The absorption and excretion of water and salts by marine teleosts. 

Amcr. J. Physiol,, 93: 480-505. 

SOLOMON, A. K., 1960. Compartmental methods of kinetic analysis. In: Mineral Metabolism 
(C. L. Comar and F. Bronner, eds.), 1A: 119-167. 



THE EFFECT OF LITHIUM AND O-IODOSOBENZOIC ACID ON THE 
EARLY DEVELOPMENT OF THE SEA URCHIN EGG 

BERNDT E. HAGSTROA1 
The Wenner-Grens Institute for Experimental Biology, University of Stockholm, Sweden 

In 1892 C. Herbst observed that the addition of lithium ions to sea water changes 
the determination of the sea urchin egg in a vegetal direction, i.e., the development 
of the endoderm becomes accentuated, whereas the differentiation of the ectoderm 
is suppressed, leading to a vegetalization of the larva. This specific effect of lithium 
has been utilized in a number of important investigations on developmental physi- 
ology; for references, see, for example, Gustafson (1954), Horstadius (1935, 1939, 
1949), Lindahl (1936, 1941), Runnstrom (1954), von Ubisch (1953). 

The vegetalizing action of lithium becomes apparent in an exogastrulation of 
the larva, i.e., there is no invagination of the gut; the ciliation of the larva is also 
reduced. These aberrant morphological characteristics are not manifested, how- 
ever, until a rather late stage of development, whereas the treatment, which gives 
rise to the changes referred to above, may be applied in immediate connection with 
fertilization. The direct effect of lithium on the larva during the actual period 
of treatment does not seem, as yet, to have been studied in detail. 

Animalization (the opposite of vegetalization), which is manifested as an in- 
crease in the development of the ectoderm and a reduced differentiation of the 
endoderm, has likewise been achieved by various treatments (cf. Lindahl, 1941 ; 
Runnstrom, 1954). These so-called animalized larvae show a strong ciliation and 
a suppressed development of the gut. It has recently been demonstrated that 
o-iodosobenzoic acid (IB A) exerts a strong animalizing action on sea urchin larvae 
(Backstrom, 1953; Runnstrom and Kriszat, 1952a, 1952b) ; the treatment with 
IBA is also applied during the early development of the larva, during the first 
hours after fertilization. 

In 1952, the author observed that lithium interferes with cell division, and also 
the movements of the chromosomes were found to be affected when lithium ions in 
concentrations of 0.05-0.06 M were present in the sea water. Owing to other in- 
vestigations in progress, these first observations on the cytological effect of lithium 
were not followed up until 1959, when the present work was started. 

MATERIAL AND METHODS 

The present study was carried out at Stazione Zoologica in Naples, where 
Paracentrotus lividus and Psammechinus microtuberculatus served as material. 
Experiments have also been made at Biologisk Stasjon, Espegrend, Norway, with 
eggs and sperm from Echinus esculcntns and Strongylocentrotus droebachiensis, 
and at Kristinebergs Zoologiska Station, Fiskebackskil, Sweden, where the gametes 
from Echinocardium cordatum and Psammechinus inilioris were used. 

55 



56 BERNDT E. HAGSTROM 

In the experiments with lithium the concentrations varied from 0.03 M to 
0.09 M. The concentrations of o-iodosobenzoic acid used varied between 5 X 10~ 5 
M and 10' 3 M. 

The experimental treatment with lithium and IBA was applied according to 
several different methods. In most experiments the active substances were added 
to batches of the egg suspension within 5-10 minutes after insemination. At differ- 
ent time intervals after the first treatment was started, new batches of eggs from 
the fertilized control were transferred to the lithium and IBA. In other series, 
lithium and IBA were added to the larvae when they were passing through a cer- 
tain cell phase, e.g. prophase or metaphase, or when they were just completing 
cell division, e.g. from the 4- to the 8-cell stage. 

In most of the experiments the treatment with the active substances was in- 
terrupted at the 64-128-cell stage, and the larvae were transferred to pure sea 
water ; in other series, the larvae were reared in the presence of lithium or IBA 
through the hatching stage. 

The cultures with cleaving larvae were counted at different time intervals after 
insemination; the percentages of cleavage given in this paper are based on counts 
of 200-250 larvae. The counts were made either in the living state or after 
fixation in Carney's fluid or in 4% formol. When the counts were made on the 
living cultures, the control and the cultures treated were in most experiments in- 
seminated successively at intervals of 5 minutes, which enables the counts to be made 
at the same point of time after insemination. 

Studies of the chromosomes and the nuclear phases were carried out after 
fixation of the larvae in Carney's fluid and staining with aceto-carmine or 
aceto-orcein. 

RESULTS 

The effect of lithium on the rate of cleavage 

When fertilized eggs were transferred to sea water containing lithium, a con- 
siderable retardation in the rate of cleavage was observed. In experiments where 
the inseminated eggs were subjected to lithium already before the sperm and egg 
nuclei had fused, this retardation became particularly evident, often resulting in a 
complete blocking of the first cell division. This was especially noticeable when 
the concentration of lithium was kept rather high, i.e., 0.06-0.09 M. 

A certain variation in sensitivity to lithium between the different species tested 
was observed and, for example, in Strongylocentrotus droebachiensis and Echinus 
csculcntns, species which develop at a rather low temperature, and consequently 
the first cleavage requires three hours or more to be completed, the blocking effect 
on the fusion of the pronuclei was more pronounced than in Paracentrotits. The 
latter species develops more rapidly and at a much higher temperature. The re- 
tardation noted at the first cleavage is evident also during the ensuing development. 

In other experiments the treatment with lithium was started at different time 
intervals after the first cell division, and also in these experiments the retarding 
effect of lithium on cleavage was striking. The sensitivity to lithium was tested 
in consecutive experiments up to the 64- to 128-cell stage, when it becomes difficult 
to observe the single cell divisions. 



EFFECT OF LITHIUM AND IBA 



57 



The effect of o-iodosobenzoic acid on the rate of cleavage 

Experiments similar to those described with lithium were carried out also with 
o-iodosobenzoic acid. Most of these experiments were made at the same time as 
those with lithium ; the same egg material and the same control were used for both 
series. 

In contrast to lithium, IBA was found to effect a general increase in the rate 
of cleavage. As in the experiments with lithium, treatment with IBA was started 
before the first cell division, and at different intervals during cleavage, i.e., at the 
2-, 4-, 8-, 16-, 32-, and 64-cell stages. The length of treatment was also varied 
as in the lithium experiments. Thus, the treatment was stopped, for instance, 
after the 16-cell stage or at the 64- to 128-cell stage. Experimental series were 
also carried out in which the treatment was not interrupted until after the larvae 
in the control had begun to hatch. 

More than 200 experiments, in which the larvae were reared over hatching, 
were made in the course of the present investigation. In many of these experi- 
ments the development was followed up to the pluteus stage. Though the experi- 
ments were carried out with six species of sea urchins, under varied and different 
experimental conditions, the results were practically uniform. The results of a 
representative experiment with eggs and sperm from Paracentrotus lividus are 
referred to in Table I. 

TABLE I 

Eggs and sperm from Paracentrotus lividus. Temperature 18.5 C. (1) Control. The eggs were 

inseminated with 10 6 spermatozoa per ml. Twenty minutes after insemination eggs 

from the control were transferred to: (2) 0.06 M Lid, and 

(3) 3.5 X 10~* M IBA 



45 minutes after insemination: 



58 minutes after insemination: 



76 minutes after insemination; 



90 minutes after insemination: 



135 minutes after insemination: 



190 minutes after insemination: 



No cleavage was observed in either (1), 
(2) or (3) 

(1) 90% uncleaved, 10% 2-cell 

(2) 100% uncleaved 

(3) 40% uncleaved, 60% 2-cell 

(1) 3% uncleaved, 97% 2-cell 

(2) 7% uncleaved, 93% 2-cell 

(3) 97% 2-cell, 3% 4-ccll 

(1) 99% 2-cell, 1%4-cell 

(2) 3% uncleaved, 97% 2-cell 

(3) 85% 2-cell, 15% 4-cell 

(1) 2% 2-cell, 50% 4-cell, 48% 8-cell 

(2) 3% 2-cell, 85% 4-cell, 12% 8-cell 

(3) 8% 4-cell, 92% 8-cell 

(1) 4% 8-cell, 96% 16-cell 

(2) 53% 8-cell, 47% 16-cell 

(3) 89% 16-cell, 11%32-cell 



58 BERNDT E. HAGSTROM 

The absolute difference in time between the control and the batches of eggs 
developing in the presence of lithium or IB A is, among other factors, dependent 
upon the species used, the concentration of the active substance and the temperature. 
The retardation obtained in the presence of 0.06 M lithium was measured to about 
70 minutes for the second cleavage of eggs from Strongylocentrotus droebachiensis 
reared at 8 C. The corresponding retardation observed for eggs from Paracentro- 
tus at 19.5 C. was recorded to about 12-15 minutes. At 18 C. this difference 
may amount to 20 minutes or more. The advance in cleavage obtained by treating 
the larvae in 7 X 10~ 4 M IB A was found to be 20-25 minutes for the second to 
fourth cleavages of Paracentrotus eggs reared at a temperature of 18.4 C. 

As has already been pointed out, many factors influence the rate of cleavage. 
Experiments of the type shown in Table I have been preferred for evaluating the 
qualitative effect of the substances tested, whereas in experiments where the abso- 
lute retarding or enhancing effect was to be studied, the cultures were fixed at the 
same stage of development and the difference in time was measured. 

The influence of lithium and IB A on the fertilisation rate 

A sensitive method of testing the positive or negative effect of substances pres- 
ent in fertilization experiments is the so-called fertilization rate method (Hagstrom 
and Hagstrom, 1954, 1959). 

Because both lithium and IB A were found to exert a strong influence on early 
development, it was thought to be of interest to investigate the action of these 
substances also on fertilization. 

Lithium and IB A were added to the egg suspensions immediately before in- 
semination, and consequently, the substances acted on both eggs and spermatozoa 
at the moment of fertilization. After 5, 10, 15, 20, etc., seconds, sodium lauryl 
sulphate was added to a final concentration of 0.001%, which instantaneously kills 
the spermatozoa without affecting the eggs (Hagstrom and Hagstrom, 1954, 1959). 
This method makes it possible to count the number of eggs fertilized 5, 10, 15, etc., 
seconds after insemination. 

The concentrations of lithium in the present experiments were 0.04 M and 
0.06 M, and the concentrations of IBA varied between 3 X 10" 4 M and 10' 3 M. 

Lithium in the concentrations tested produced no effect on the fertilization rate. 
In the experiments with IBA there was also no clear difference recorded when 
compared with the control series. In some cases, however, there was a slight im- 
provement with IBA in concentrations of about 10~ 3 M. No deleterious effects on 
the processes at fertilization due to either lithium or IBA were observed. This is 
of importance for the evaluation of the cleavage rate experiments. 

The influence of lithium and IBA on hatching 

The first developmental period ends when, at hatching, the sea urchin larva 
breaks the fertilization membrane and can swim freely. The series of events lead- 
ing up to hatching makes it possible to record a number of objective observations 
on the rate of development. 

As previously mentioned, the duration of treatment varied, and in some experi- 
ments it was interrupted at a relatively early stage, i.e., in the 64- to 128-cell stage, 



EFFECT OF LITHIUM AND IBA 59 

whereas in other experiments the active substances were present from a few minutes 
after insemination until the larvae started to hatch in the control. Many variations 
were applied within this general scheme of treatment. 

In experiments where the larvae were reared in lithium or IBA from insemina- 
nation until the larvae of the control began to hatch, these substances were found 
definitely to inhibit hatching. It is known that IBA prevents hatching, and that the 
larvae may then develop inside the fertilization membrane (Runnstrom and Krisxat, 
1952b). The present experiments showed that lithium also tended to arrest hatch- 
ing. There is, however, a considerable difference between the action of lithium 
and that of IBA. Lithium not only delays or completely inhibits hatching, but 
also prevents the formation of cilia and the movements of the larva inside the intact 
membrane. In IBA the actual rupture of the membrane is obviously impeded, 
whereas the ciliation of the larva is normally or even more than normally developed. 
Consequently, in the presence of IBA the larvae acquire a high degree of rotatory 
activity inside the membranes. This is also in agreement with the fact that the 
larvae develop more rapidly in IBA; accordingly, the viability of the larvae seems 
to be high. A typical experiment is referred to in Table II. 

In certain experiments with high concentrations of lithium and IBA, hatching 
was completely arrested, especially when the slowly developing Norwegian species 
were used. 

Experiments in which the active substances were removed at a relatively early 
stage of development were also carried out. Qualitatively, the same results were 
gained as in the experiments mentioned above; lithium gives rise to a general de- 
crease in the rate of development, whereas IBA promotes development, with the 
exception that IBA affects hatching. 

The negative effect exerted by IBA on hatching must be ascribed to the fact 
that the rupture of the fertilization membrane is arrested, and that this process 
is not correlated with any retardation of the development in general. 

Morphological observations on the cleavage stages 

The effect of lithium and IBA on the nuclei and chromosomes appeared to 
be of special importance. Observations were made with the phase contrast micro- 
scope on vital material, and, moreover, about 1500 fixations have been prepared 
during the course of this investigation. The results of the cytological studies will 
be reported elsewhere. The present paper will deal only with observations on the 
cleavage pattern of the young larva. 

With IBA, cleavage seems to be normal though accelerated. The cleavage 
furrows are deep, but the blastomeres remain well attached to each other. 

The effects observed after treatment with lithium were considerably more com- 
plex. Lithium was found to induce a very clear separation of the blastomeres. This 
deterioration in the contacts between the cells of the cleaving blastulae is likely to 
affect the transport mechanism within the larva and the exchange of metabolites 
between the different regions of the embryo, thus interfering with a mechanism 
which is undoubtedly of the utmost importance to the ensuing development. When 
the micromeres form, they seem, however, to retain intercellular contact with each 
other and with the macromeres. This may be due to the fact that the micromeres 
have a small volume but a relatively large surface area as compared with the macro- 



BERNDT E. HAGSTROM 

TABLE II 

Paracentrotus lividus, eggs from one female. Temperature 18 C. (1) Control. The eggs were 

inseminated with 10 s spermatozoa per ml. Twelve minutes after insemination eggs 

from the control were transferred to: (2) 0.06 M LiCl, and 

(3) 7X1" M IBA 

9 hours 50 minutes after insemination: 

(1) A few ciliated larvae rotating inside the membranes. No hatched larvae. 

(2) No ciliation or hatching. 

(3) Most larvae ciliated and moving inside the membranes. No hatched larvae. 

10 hours 15 minutes after insemination: 

(1) Increased ciliation and movements. No hatching. 

(2) No ciliation. 

(3) Vigorous movements, stronger than in (1). No hatching. 

10 hours 40 minutes after insemination: 

(1) A few hatched larvae. 

(2) No ciliation, no movements or hatching. 

(3) Strong rotation inside the membranes. No hatching. 

11 hours 10 minutes after insemination: 

(1) About 50% hatched larvae. 

(2) A few larvae with weak ciliation. 

(3) Vigorous movements, a few hatched larvae. 

11 hours 55 minutes after insemination: 

(1) About 80% hatched larvae. 

(2) About 25% with weak rotating movements. No hatching. 

(3) Vigorous movements, about 40% hatched larvae. 

12 hours 20 minutes after insemination: 

(1) 95% hatched larvae. 

(2) No hatching. 

(3) About 60% hatched larvae. 

12 hours 55 minutes after insemination: 

(1) 100% hatched larvae. 

(2) No hatching, weak rotation inside the membranes. 

(3) About 80% ruptured membranes. 

14 hours after insemination: 

(2) About 30% ruptured membranes. The ciliation was still very weak. 

(3) 90% hatched larvae. The larvae showed higher mobility than in the control. 

14 hours 55 minutes after insemination: 

(2) 80% with ruptured membranes. Low mobility. 

(3) 100% hatched larvae. 

16 hours 20 minutes after insemination: 

(2) 90% hatched larvae swimming near the bottom with low mobility. 

(3) The primary mesenchyme was well developed in (3) but was not present in the 

control or in (2). 

meres. The position of the micromeres, squeezed in between the macromeres, may 
also facilitate intercellular contacts between these two types of cells. 

As pointed out previously, lithium interferes with the rate of cleavage, and 



EFFECT OF LITHIUM AND I1!A 



61 



I \iiLi-: III 

Paracentrotus lividus, eggs from one female. Temperature 18 C. (1) Control. The eggs were 

inseminated with 10 6 spermatozoa per nil. Eggs from the control were transferred 

into 0.05 M LiCl after: (2) 5 minutes; (3) 215 minutes 



120 minutes after insemination 



140 minutes after insemination: 



165 minutes after insemination 



200 minutes after insemination: 



230 minutes after insemination: 



260 minutes after insemination : 



290 minutes after insemination 



320 minutes after insemination: 



365 minutes after insemination: 



(1) 39% uncleaved, 61% 2-cell 

(2) 94% uncleaved, 6% 2-cell 

(1) 1% uncleaved, 99% 2-cell 

(2) 52% uncleaved, 48% 2-ccll 



(1) 78% 2-cell, 22%4-cell 

(2) 5% uncleaved, 95% 2-cell 

(1) 27% 2-cell, 70% 4-cell, 3% 8-cell 

(2) 87% 2-cell, 13% 4-cell 



(1) 64% 4-cell, 33% 8-cell, 3% 16-cell 

(2) 48% 2-cell, 48% 4-cell, 4% 8-cell 

(3) 65% 4-cell, 33% 8-cell, 2% 16-cell 

(1) 14% 4-cell, 68% 8-cell, 18% 16-cell 

(2) 15% 2-cell, 74% 4-cell, 11% 8-cell 

(3) 3% 2-cell, 20% 4-cell, 66% 8-ce!l, 11% 16-cell 

(1) 56% 8-cell, 44% 16-cell 

(2) 18% 2-cell, 42% 4-cell, 33% 8-cell, 7% 12-cell 

(3) 3% 4-cell, 56% 8-cell, 2% 12-cell, 39% 16-cell 

(1) 26% 8-cell, 71% 16-cell, 3Vo 32-cell 

(2) 15% 2-cell, 29% 4-cell, 10% 8-cell, 40% 12-cell, 
6% 16-cell 

(3) 2% 4-cell, 40 r ;, 8-cell, 5% 12-cell, 53% 16-cell 

(1) 6% 8-cell, 62% 16-cell, 32% 32-cell 

(2) 2% 2-cell, 2% 4-cell, 6% 8-cell, 28% 12-cdl, 
62% 16-cell 

(3) 7', 8-cell, 17' , 12-cell, 62% 16-cell, 14% 32-cell 



within the larvae there appears to be a certain gradation in response to lithium ; the 
different morphological regions of the embryo seem to be differently affected. 

When the larvae have reached the 8-cell stage, and the process of cell division 
begins, which leads to the 16-cell stage, some of the embryos formed intermediate 
12-cell stages (italicized in Table III) instead of the normal 16-cell stages. The 
cleavages in the four animal cells of the larva became temporarily inhibited when 
influenced by lithium, whereas the cleavages of the vegetal cells were not arrested 
to the same extent. The blocking of the cleavages in the animal region of the larvae 
is incomplete and temporary. As a consequence of the delayed formation of the 
presumptive mesomeres, the balance between the animal and the vegetal halves of 
the embryo becomes disturbed. The delayed cleavages in the animal half are also 



62 BERNDT E. HAGSTROM 

evident at the formation of the 32-, 64- and 128-cell stages, when larvae with 
reduced numbers of animal cells occur frequently. A typical experiment is shown 
in Table III. 

DISCUSSION 

The results obtained in the present investigation indicate that the effects, which 
are recorded on cleavage after treatment with IBA and lithium, are correlated with 
the "animalization" and "vegetalization" observed during the ensuing development 
of the larvae. The enhanced or delayed rate of cleavage is not necessarily the 
direct cause of the animalizing or vegetalizing effect, but, in the author's opinion, 
the alteration in the rate of cleavage reflects the primary changes in the cells of the 
young larva, which result in the secondary events that occur during the later 
development. 

The present observations indicate that the nuclei and chromosomes are affected 
by lithium, which interferes also with cell division. It has been shown that lithium 
ions cause a deficiency in the nucleoprotein synthesis in Xenopus embryos (Thoma- 
son, 1957), and this finding has probably some bearing on the present results. 

The disruptive action of lithium on the cell contacts appears to be of special 
importance, because this is likely to render difficult the exchange of metabolites 
within the embryo. A phenomenon, which cannot be fully explained at present, is 
the high proportion of 12-cell stages (instead of 16-cell stages) observed after 
treatment with lithium. The fact that the reduced number of cells is dependent 
upon temporarily inhibited cleavages in the presumptive mesomeres indicates that 
the equilibrium within the larva is disturbed, and that the vegetal part of the 
embryo obtains a certain lead in development over the animal part. It may also 
indicate that the animal cells are more sensitive than the vegetal to exposure to 
lithium. The cell divisions are not equal, however (cf. Horstadius, 1935), in the 
four animal and the four vegetal cells of the 8-cell stage, and consequently, they 
are not directly comparable. This indicates that the apparent differences in 
sensitivity to lithium may as readily be ascribed to the different orientation of the 
mitotic spindles in the animal and the vegetal cells, which may per se dispose the 
cells to respond differently to lithium. As has been previously mentioned, a 
similar effect of lithium was observed when the larvae were treated at the 16- 
to 32-cell stage, at the 32- to 64-cell stage, and at the 64- to 128-cell stage ; larvae 
with less than the normal number of cells were also frequently encountered here. 

Whether this disturbance of the cleavage pattern is to be attributed to a real 
difference in sensitivity between the animal and the vegetal cells of the embryo 
may, at present, be left an open question. At the cleavage stages between 16 and 
128 cells, the cells of the animal and the vegetal halves are rather unequal. These 
cells have not the same surface areas, and the ratio, surface area/volume, is also 
different in the animal half from that in the vegetal. If we assume that the uptake 
and the action of lithium are correlated with the surface area exposed to the active 
ions, it may be justifiable to conclude that the deleterious effect on the animal part 
of the larva is visible on account of the larger surface area exposed to lithium. 
However, if there is any graded response to lithium in the different parts of the 
embryo, it is likely to be of a temporary nature. The elaborate experiments carried 
out by Lindahl and Holter (1940) point in the same direction. Their results 



EFFECT OF LITHIUM AND IBA 63 

showed that isolated animal and vegetal halves have the same oxygen consumption. 
Furthermore, it was demonstrated that treatment with lithium evokes the same 
inhibition of respiration in both animal and vegetal cells (Lindahl and Holter, 1940). 

It is evident that lithium causes a general slowing down of early development. 
Lindahl (1936, 1941) showed that the oxygen consumption of larvae reared in 
the presence of lithium is considerably lower than that for larvae which develop in 
pure sea water ; this is in agreement with the results reported here. The lithium 
effect becomes especially marked at the time of hatching, when the ciliation of the 
larvae treated is either reduced or entirely lacking. In some experiments, hatch- 
ing took place before any ciliation of the larvae had developed, which may indi- 
cate that the production of the hatching enzyme (cf. Lundblad, 1954) is less 
affected by lithium than are the processes leading to ciliation. 

The results obtained with IBA show that the action of this substance enhances 
cleavage. The ciliation of the larva is extremely well developed, and the rotatory 
movements inside the membrane often begin earlier in a treated batch of larvae 
than in the corresponding control. The actual rupture of the membrane is, how- 
ever, delayed or even prevented. It was previously shown (Backstrom, 1955 ) that 
IBA does not interfere with the respiration of the larva up to the hatching stage, 
which is in accordance with the present results. IBA has not, as yet, been ob- 
served to cause any unbalanced increase in the cleavage of, for example, the animal 
half of the larva, but all cells within the embryo seem to be subject to the same 
effect that promotes cleavage. Though a number of substances and treatments 
have been found to induce animalization of the sea urchin larva, no common de- 
nominator for their physiological action has so far been discovered. It is therefore 
of interest to note that trypsin, which also induces an animalizing effect (Horsta- 
dius, 1949, 1953), has recently been shown to bring about an accelerated rate of 
cleavage comparable with that resulting from treatment with IBA (Hagstrom and 
Lonning, 1962). 



My sincere thanks are due to Professor S. Horstadius, Uppsala, for valuable 
suggestions and for reading the manuscript. 

The present investigation was supported by grants from the Swedish Natural 
Science Research Council which are gratefully acknowledged. 

SUMMARY 

Under varied experimental conditions young sea urchin larvae were subjected to 
the action of solutions of o-iodosobenzoic acid and LiCl in sea water. Cleavage and 
early development were found to be advanced by IBA, and retarded by lithium. 
The fusion of the pronuclei was strongly inhibited by lithium, which tends also to 
separate the blastomeres. The rate of fertilization was not appreciably influenced 
by either IBA or lithium in the concentrations tested. 

LITERATURE CITED 

BACKSTROM, S., 1953. Studies on the animalizing action of iodosobenzoic acid in the sea urchin 
development. Arkiv Zool., 4: 485-491. 



64 BERNDT E. HAGSTROM 

BACKSTROM, S., 1955. The effect of iodosobenzoic acid on the respiration of the developing sea 

urchin egg. Arkiv Zool., 7: 573-578. 
GUSTAFSON, T., 1954. Enzymatic aspects of embryonic differentiation. Intern. Rev. CytoL, 3: 

277-327. 
HAGSTROM, B., AND B. HAGSTROM, 1954. A method of determining the fertilization rate in sea 

urchins. Exp. Cell Research, 6: 479-484. 
HAGSTROM, B. E., AND B. HAGSTROM, 1959. The effect of decreased and increased temperature 

on fertilization. Exp. Cell Research, 16: 174-183. 
HAGSTROM, B. E., AND S. LONNING, 1962. The effect of trypsin on the early development of the 

sea urchin egg. Arkiv Zool. in press. 
HERBST, C., 1892. Experimentelle Untersuchungen uber den Einfluss der veranderten chemischen 

Zusammensetzung des umgebenden Mediums auf die Entwicklung der Tiere. Zeitschr. 

wiss. Zool., 55: 446-518. 
HORSTADIUS, S., 1935. Uber die Determination im Verlaufe der Eiachse bei Seeigeln. Pubbl. 

Stats. Zool. Napoli, 14: 251-429. 
HORSTADIUS, S., 1939. The mechanics of sea urchin development, studied by operative methods. 

Biol. Rev., 14: 132-179. 
HORSTADIUS, S., 1949. Experimental researches on the developmental physiology of the sea 

urchin. Pubbl. Stas. Zool. Napoli, 21 : Suppl. 131-172. 
HORSTADIUS, S., 1953. Vegetalization of the sea-urchin egg by dinitrophenol and animalization 

by trypsin and ficin. /. Embryol. Exp. Morph., 1 : 327-348. 
LINDAHL, P. E., 1936. Zur Kenntnis der physiologischen Grundlagen der Determination im 

Seeigelkeim. Ada Zool. (Stockholm), 17: 179-365. 
LINDAHL, P. E., 1941. Physiologische Probleme der Entwicklung und Formbildung des 

Seeigelkeimes. Natunviss., 29: 673-685. 
LINDAHL, P. E., AND H. HOLTER, 1940. Beitrage zur enzymatischen Histochemie. Die Atmung 

animaler und vegetativer Keimhalften von Paracentrotus lividus. C. R. Trav. Lab. 

Carlsberg, Ser. Chim., 23: 257-288. 
LUNDBLAD, G., 1954. Proteolytic activity in sea urchin gametes IV. Further investigations of 

the proteolytic enzymes of the egg. Arkiv Kemi, 7: 127-157. 
RUNNSTROM, J., 1954. Die Analyse der primaren Differenzierungsvorgange im Seeigelkeim. 

Verhandl. deut. Zool. Ges. Tubingen, (3) : 32-68. 
RUNNSTROM, J., AND G. KRISZAT, 1952a. Animalizing action of iodosobenzoic acid in the sea 

urchin development. Exp. Cell Research, 3: 497^99. 
RUNNSTROM, J., AND G. KRISZAT, 1952b. The action of SH-reagents on the activation process 

in the sea urchin egg. Arkiv Zool., 4: 165-185. 
THOMASON, D., 1957. Effect of lithium ions on the penetration of phosphate into Xenopus 

embryos. Nature, 179: 823. 
VON UBISCH, L., 1953. Entwicklungsprobleme. Gustav Fischer Verlag, Jena. 




OOGENESIS AND RADIOSENSITIVITY IN COCHLIOMYIA 
HOMINIVORAX (DIPTERA: CALLIPHORIDAE) 

LEO E. LACHANCE AND SARAH B. BRUNS 

Entomology Research Division, Agric. Res. Serv., U.S.D.A., 
Livestock Insects Investigations, Kerrville, Texas 

Ionizing radiation and some chemicals are known to render insects sterile. 
Sterility may be achieved in two ways, by a treatment that inhibits the formation or 
development of mature sperm or ova (fecundity), or by a treatment that does not 
deter the production of these cells per se, but induces dominant lethal changes in 
their hereditary material, rendering them incapable of sustaining embryonic growth 
or causing death in the post-embryonic stages (fertility). The end result sterility 
is the same, but the processes involved in producing it are quite different. Many 
studies have investigated the reproductive system of both males and females and 
their differences in radiosensitivity with respect to the induction of dominant lethals, 
but inhibition of the growth of the reproductive organs has not been so widely 
studied. 

In many species, growth of the ovaries is difficult to measure, especially in or- 
ganisms that produce only a few mature gametes at a time. In some insects, how- 
ever, the females produce a large number of gametes synchronously (egg masses). 
When many cells are undergoing growth and maturation simultaneously in many 
ovarioles, the reproductive organs attain considerable size at sexual maturity and 
the effect of a treatment can be easily measured. 

Some aspects of the development of the female reproductive system in the screw- 
worm fly, Cochliomyia hominivorax (Cqrl.), have been reported, in connection 
with work on radiation-induced dominant lethal mutations in developing oocytes 
(LaChance and Leverich, 1962). The present studies are concerned with the 
inhibition of ovarian growth by radiation, and with the morphological and cytologi- 
cal changes that lead to sterility characterized by infecundity. Particular emphasis 
is given here to the early growth stages, which are especially radiosensitive, and to 
the comparative radiosensitivity of the stages in oogenesis. 

MATERIALS AND METHODS 

Cochliomyia hominivorax is an obligate parasite of warm-blooded animals. 
The biology and laboratory-rearing procedures for this insect have been discussed 
in several publications (see Bushland, 1960). The insects used in the experi- 
ments reported here were of the laboratory strain, reared on artificial medium. In 
this rearing technique, the larvae pass through the three instars in the artificial 
medium and pupate in sand. The length of the pupal period is approximately 
7-7\ days at 80 F. The adults were reared at 80 F. and fed honey and water, 
a diet that supports life and ovarian growth very well. Laboratory-reared adults 
mate when approximately two days old. The females are gravid at 5-6 days of age, 

65 



LEO E. LACHANCE AND SARAH B. BRUNS 

and will produce an egg mass of more than 200 eggs when oviposition is induced 
by presenting the females with a small piece of lean meat and keeping them at a 
temperature of 90-96 F. for a few hours. A second egg mass may be produced 
3-5 days later. 

The radiation exposures were performed in air in a Co 00 radiation source 
(Jefferson, 1960) at a dose rate of 642-683 roentgens per minute. Calibration of 
the unit was accurate to 6%. 

In the studies of ovarian growth, both ovaries were dissected from individual 
females ; 30 females were examined for each treatment group. The dimensions of 
each ovary were measured with an ocular micrometer in the eyepiece of a dissecting 
microscope (17 <). One measurement of height and two of diameter (at right 
angles to each other) were made. Estimation of ovarian volume is somewhat 
difficult, since the organ does not resemble any geometric form precisely. The 
several formulae used to calculate the size of a rounded geometric object have 
certain values that are constant. However, since the measurements made in these 
experiments were not intended to establish an exact volume but to reduce the 
observations to a single number that could serve as a basis for comparison be- 
tween groups, the volume of the ovaries was approximated by multiplying the 
three measurements and correcting for the magnification. 

In the studies of ovarian growth in Cochlioinyia, temperature was repeatedly 
observed to affect size : Females of equal age reared at 80 F. had larger ovaries 
than those reared at 73 F., and sexual maturity was reached in 5 days at the 
higher temperature whereas 7 days was required at the lower temperature. For 
this reason, rearing temperatures in these experiments were all closelv controlled 
at 80 F. 

For the cytological studies, whole mounts of ovarioles were made from females 
of various ages and stained on a microscope slide with the Feulgen procedure. In 
all studies, corroborative evidence was gained by examining freshly prepared 
mounts in insect saline with a Zeiss phase-contrast microscope at 1600 X. The 
staining procedure causes considerable shrinkage of the ovarioles, so those examined 
in wet mounts were always somewhat larger than the Feulgen-stained preparations. 

RESULTS AND DISCUSSION OF INDIVIDUAL EXPERIMENTS 
1. Growth of the ovaries in normal females 

Data on the measurements of the ovaries in normal females of various ages are 
summarized in Table I. Newly emerged females had immature ovaries with 
volumes ranging from 0.068 to 0.254 mm 3 . The size of the ovary doubled between 
the first and second day of adult life and increased by more than five-fold between 
the second and third day of adult life. Growth then continued somewhat more 
slowly until the mature ovaries had attained a size of more than 7 mm 3 . Thus, 
from emergence until sexual maturity, the size of the ovary increased approximately 
60-fold. 

The growth of the ovary is the manifestation of the growth processes occurring 
synchronously in the more than 100 ovarioles that comprise each ovary. A cyto- 
logical study of the ovarioles was conducted to determine the sequence of events in 
normal oogenesis, so that a basis could be established for evaluating the effects of 



OOGENES1S AND RADIOSENSITIVITY 67 

TABLE I 

Growth of ovaries in normal Cochliomyia hominivorax. (Each number represents the 
mean from 30 females the standard error of the mean.) 

Mean volume of ovaries Si (mm. 3 ) 
Age of female 



(days 2 hours) Right 

0-2 hours 0.1423 0.0225 0.1273 0.0015 

1 0.2022 0.0095 0.2002 0.002S 

2 0.4094 0.0470 0.3952 0.0362 

3 2.1064 0.204 2.2446 0.234 

4 5.3841 0.328 5.0761 0.347 

5 7.3885 0.274 7.3921 0.274 

radiation or other damaging agents on the growth of the reprodvictive organs 
and on other factors affecting female sterility. 

2. Cytology of normal ovarioles 

The ovary in Cochliomyia consists of 100-150 ovarioles, each of which is en- 
closed in an epithelial sheath and produces one mature egg to be deposited in each 
egg mass. Oocyte development is synchronous in all ovarioles. Scale drawings 
showing the normal developmental sequence in an ovariole are presented in Figures 
8 and 9. The cytological studies showed the following pattern of oogenesis corre- 
lated with the age of the female. 

Pupae 4 and 5 days old. The ovary is small and immature. Each ovariole is 
composed of a germarium filled with oogonial cells (Fig. 8A and Fig. 1). These 
cells are mitotically active, and nests of four or eight cells in division are occasionally 
seen in the anterior and middle portion of the germarium. There is no evidence 
of nurse cell formation. Examination of ovarioles from 4- and 5-day-old pupae, in 
Feulgen-stained preparations or in phase-contrast studies of fresh whole mounts 
in saline, did not show any differentiation of the germarial contents. 

Pupae 6 days old. The ovariole is still composed of a single germarium, sur- 
rounded by an envelope of epithelial cells (Fig. SB and Fig. 2). Mitotic divi- 
sions are now relatively rare. In some of the ovarioles the nuclei in the posterior 
portion of the germarium have enlarged very slightly, and the chromatin material 
has become slightly more diffuse. In some ovarioles a faint line of separation is 
seen between the anterior and posterior portion of the germarium (Fig. 2) ; this is 
the first indication of the first egg chamber being formed. 

Pupae 7 days old. A cyst has formed in the posterior portion of each ovariole, 
indicating formation of the first egg chamber. In some ovarioles dissected from 
pupae almost ready to emerge, an indentation has formed that separates and 
''pinches off" the first egg chamber from the germarium (Figs. 3 and 8C). This 
newly formed chamber contains 16 cells, 15 of which are large and distinct nurse 
cells and the other, the smaller oocyte. The 15 nurse cells have not differentiated 
noticeably, except that the nuclei contain diffuse chromatin material. The germar- 
ium occasionally contains a smaller second cyst, in the posterior region, that is 
destined to become the second egg chamber (Fig. 3). This structure was visible in 
phase contrast ; in Feulgen-stained preparations, changes in nuclear morphology 
were not observable in the remaining germarial contents. 



68 



LEO E. LACHANCE AND SARAH B. BRUNS 




1 




FIGURES 1-2. Phase contrast, unfixed tissue in saline, 640 X. Figure 1. Ovariole from 
untreated 5-day-old pupa. Figure 2. Ovariole from untreated 6-day-old pupa ; note furrow 
forming in posterior region of germarium. 

Presumably the 16 cells of an egg chamber derive from four consecutive divi- 
sions of an oogonial cell, maintain a positional relationship to one another, and are 
subsequently incorporated into an egg chamber or cyst, but in Cochliomyia there 



OOGENESIS AND RADIOSENS1TIVITY 69 

was no indication that these cells are enclosed in a cyst until the pupae are 61-7 
days old. Very likely the daughter cells are interconnected by fine protoplasmic 
strands, but these were not demonstrable with the techniques used in these studies. 
Hirschler (1955) has presented an excellent study of the early positional relation- 
ship of the cells that will subsequently develop into an egg-nurse cell complex in 
insects. He states (p. 78), "After a given number of divisions, the egg remains 
united to the conceivably greatest number of nurse-cells by means of cell bridges 
functioning as nutritive ducts. As far as nutrition is concerned, the egg lies in the 
best possible situation." The studies of King (1960) also indicate that these 
daughter cells remain connected by intercellular cytoplasmic bridges and are sub- 
sequently incorporated into an egg chamber. Fawcett ct al. (1959) have studied 
the intercellular bridges that exist between daughter cells derived from an original 
parent cell in the spermatids of many species, and suggest that protoplasmic con- 
tinuity between these cells is the basis for their synchronous development. It is 
quite possible that the synchronous development of ovarian nurse cells is also 
related to their intercellular connections. 

Adult females 0-4 hours old. The 15 nurse cells in the first egg chamber are 
greatly enlarged (Fig. 8D). The chromosomes in the nurse cell nuclei are 
thickened and stain deeply with Feulgen stain (Fig. 4A and 4B). Formation of 
polytene chromosomes in ovarian nurse cells is quite common in Diptera (Stalker, 
1954; Bier, 1960). In Cochliomyia the reduplicated chromonemal strands remain 
in close enough association to appear as greatly thickened chromosomes (Fig. 4A), 
but not so close as to give a very definite banded appearance as in some other 
species (Stalker, 1954; Bier, I960). 

Adult females 4-24 hours old. The major changes in the first egg chamber 
take place in the nurse cell nuclei, which continue to enlarge progressively, due to 
the replication of chromosomal material. This process results in the formation of 
polytene chromosomes. Chromonemal reproduction in the nurse cells is followed 
by a complete separation of the reproduced elements. The separation does not 
involve a regular dicentric anaphase, but merely a complete dissociation of the 
numerous strands or a general "falling apart," resulting in a mass of chromatin 
fibrils that completely fill the nuclear volume (Fig. 5). The entire process takes 
place within an intact nuclear membrane and may, according to the terminology of 
Lorz (1947) and others (see Painter and Reindorp, 1939; Painter, 1959), be 
termed endomitosis. This endomitotic process has been described in great detail 
by Painter and Reindorp (1939) and King (1960) for Drosophila, and by Bier 
(1960) for Calliphora. In Cochliomyia it is very similar, except that it takes place 
simultaneously in hundreds of ovarioles. Jacob and Sirlin (1959) have reported 
that in Drosophila the most posterior nurse cells undergo one more duplication than 
the more anterior cells, and become larger and more active. The posterior nurse 
cells in Cochliomyia are also distinctly larger (Fig. 5), and very likely undergo 
more replications of chromosomal material than do the anterior nurse cell nuclei. 

In these studies, only rarely were the polytene chromosomes of the nurse 
cells observed to dissociate before the females were 8 hours old. In ovarioles 
from females 8-16 hours old, the thickened polytene chromosomes had dissociated 
into Feulgen-positive chromatin threads in approximately half of the nurse cells. 

Adult females 24 hours old. The endomitotic process is completed during the 
first day of adult life. After 24 hours all the nurse cell nuclei are filled with loosely 



70 



LEO E. LACHANCE AND SARAH B. BRUNS 




OOGENESIS AND RADIOSENSITIVITY 71 

associated Feulgen-positive chromatin threads (Fig. 5i. The nurse cells have 
now enlarged to their greatest size. The ovariole consists of an enlarged first egg 
chamber, the second chamber is beginning to form in some mnrioles, and the 
germarium is filled with oogonial cells (Fig. 9A). In contrast, the oocyte nucleus 
in newly emerged females is smaller and stains very lightly with Feulgen stain 
throughout the first day. In these studies the nucleolus did not stain, but it was 
clearly visible in phase-contrast examination. 

Adult females -/<V hours old. The first egg chamber has enlarged considerably, 
so that the nurse cells are not so closely packed, but more dispersed in the chamber. 
The oocyte nucleus is very small and spherical, and is located in the posterior region 
of the egg chamber. The second egg chamber has now clearly formed in all 
ovarioles. The nurse cells in the first egg chamber are extruding a fine granular 
material which stains faintly with Feulgen ; this material is beginning to fill the 
egg chamber. 

Adult females 3 days old. Between the second and third day after emergence, 
the first egg chamber in each ovariole more than doubles in size (see Table I). 
The nurse cells are localized at the anterior end of the chamber (Fig. 9B and C). 
The oocyte has grown noticeably and is beginning to elongate ; the oocyte nucleus 
is located in the ooplasm very near the nurse cells and is in prophase I of the first 
meiotic division (LaChance and Leverich, 1962). The second egg chamber has 
also enlarged and contains nurse cells undergoing endomitotic replications. Some 
of the nurse cells in the second egg chamber have completed the process of endo- 
mitotic growth and dissociated chromosomal material is beginning to fill the 
nucleus. 

Adult females 4 days old. The mature ovum is now almost fully formed. The 
nurse cells, when present, stain much more deeply and appear to darken and 
disintegrate. Most nurse cells in the upper portion of the egg follicle have disap- 
peared (Fig. 9D). The disappearance of the nurse cells from the first egg chamber 
is correlated with the passage of the oocyte nucleus from prophase I to metaphase 
I of the first meiotic division (LaChance and Leverich, 1962). 

Adult females 5 days old. Each ovariole contains a mature ovum ready for 
oviposition ( Fig. 7a ) . The nurse cells have completely disappeared and the oocyte 
nucleus is now in anaphase I of the first meiotic division (LaChance and Leverich, 
1962). All nurse cells in the second egg chamber have completed endomitotic 
growth and contain nuclei filled with fine Feulgen-positive threads existing in great 
multiplicity. A very small third egg chamber has just formed from the germarium. 

Thus, at the time when the first egg mass is deposited, each ovariole contains 
a well developed second egg chamber, which will then repeat the process described 
above and form a mature ovum to be deposited in the second egg mass. At 5 days 
of age, growth in the ovarioles is arrested until deposition of the first egg mass ; 
cytological preparations of ovarioles from females 7-9 days of age that have not 
oviposited closely resemble those of females 5 days of age. 

FIGURE 3. Phase contrast, unfixed tissue in saline, 640 X. Ovariole from untreated pupa 
7-7-2 days old; first egg chamber well formed (upper right), germarium containing a second cyst 
(lower left). 

FIGURE 4. Feulgen-stained whole mounts, 800 X. Egg chambers in ovarioles from untreated 
adults 0-4 hours old. (A) Two adjoining egg chambers, showing polytene nurse cell chromo- 
somes (lower focal level). (B) One egg chamber, showing polytene nurse cell chromosomes 
( upper focal level ) . 



72 



LEO E. LACHANCE AND SARAH B. BRUNS 




FIGURE 5. Feulgen-stained whole mount, 512 X. Ovariole from untreated 28-hour-old 
adult; note 15 nurse cell nuclei containing dissociated chromatin fibrils, oocyte nucleus very 
faintly stained (lower left of first egg chamber). 








B 



FIGURE 6. Feulgen-stained whole mounts, 80 X. Ovarioles from 2i-day-old adults. (A) 
Control ; shows large first egg chamber with 15 enlarged nurse cells, second egg chamber, and 
germarium. (B) Treated with 2000 r as 5-day-old pupa; note absence of second egg chamber. 
(C) Treated with 2000 r as adult 0-4 hours old. 



OOGEXESIS AND RADIOSENS1T IYITY 



73 










, 



*s + 






I 



B 



FIGURE 7. Feulgen-stained whole mounts, 80 X. Ovarioles from 4^-day-old adults. (A) 
Control. (B) Treated with 4000 r as S-day-old pupa; note typical undeveloped first and second 
egg chambers and atrophied germarium. (C) Treated with 4000 r as adult 0-4 hours old; note 
typical malformed egg chambers and germarium. 



3. Effects of irradiation on ovarian growth 

With the data obtained on ovarian measurements and the cytological studies of 
ovarioles from normal females of various ages, it was possible to study the effects 
of gamma radiation treatments, and to determine which ages presented the re- 
productive system in stages of highest radiosensitivity or resistance. By observing 
the inhibition of ovarian growth induced in females of various ages by a given dose 
of radiation, it was possible to compare the sensitivity of oogonial cells, of egg 
chambers in which the nurse cells were undergoing endomitotic growth, and of 



74 



LEO E. LACHANCE AND SARAH B. BRUXS 




S 



o c 




FIGURE 8. Normal ovarian development in Cochliomyia hominivorax. Each figure repre- 
sents a single ovariole ; drawings prepared from unfixed whole mounts examined in phase contrast 
(enlarged 400 X). (A) From pupa 4i-5 days old. (B) From pupa 6 days old. (C) From pupa 
7^-8 days old. (D) From adult 0-4 hours old. (o.s. = ovariole sheath; g. = germarium ; 
e.c. = first egg chamber ; o.c. = oogonial cells ; n.c. = nurse cells.) 



older egg chambers in which endomitosis was completed but vitellogenesis was in 
progress. The results of these comparisons are presented in Table II. 

The early experiments of Bushland and Hopkins (1953) showed that larvae 
and young pupae were severely injured by irradiation and emergence was reduced. 
However, 6-day-old pupae could be sterilized by a dose of 5000 r. When pupae 
were treated at this level, most of the adult females did not produce eggs and the 
few that did produced a very few eggs, which did not hatch ; the adult males 
produced motile sperm that contained dominant lethals. In further work on 6-day- 
old pupae (Baumhover ct a!., 1955), it was found that a dose of 7500 r was 
required to prevent egg production completely. The studies in Table II show that 
the process of ovarian growth can be inhibited at other times in the life cycle with 
smaller doses of radiation. Thus, although between 4000 and 5000 r were required 
to inhibit growth of the ovaries when 5-day-old pupae were irradiated, newly 
emerged females were much more radiosensitive : a dose of 2000 r reduced ovarian 
growth by half, and almost complete inhibition of ovarian growth was obtained 
with 4000 r. However, when two-day-old females were given even higher doses 
(up to 8000 r), ovarian growth and egg production were not greatly affected. It 
should be noted that, although irradiation of older females does not greatly inhibit 



OOGENESIS AND RADIOSENSITIVITY 



75 




e.c. 




FIGURE 9. Normal ovarian development in Cochliomyia hominivorax. Each figure repre- 
sents a single ovariole ; drawings prepared from stained material. (A) From adult 28 hours 
old (384 X). (B) From adult 2i days old (60 X). (C) From adult 3i days old (60X). 
(D) From adult 4 days old (60X). (n.c. = nurse cells; o.n. = oocyte nucleus; e.c. = second 
egg chamber ; g = germarium.) 

ovarian growth, sterility can still be achieved by inducing dominant lethal changes 
in the formed or forming oocytes so that, even though eggs are produced, embryos 
will not develop (LaChance and Lever ich, 1962). 



TABLE II 

Effects of gamma radiation on ovarian growth in Cochliomyia hominivorax. (Each number 
represents the mean from 30 5 -day-old females the standard error of the mean.) 



Mean volume of ovaries s x (mm. 3 ) from females treated as: 



Radiation 
dose fr) 


5-day-old pupae 


Adults 0-4 hours old 


Adults 48-52 hours old 


Right 


Left 


Right 


Left 


Right 


Left 


Control 
2000 
4000 
5000 
6000 
8000 


8.14100.277 
8. 1102 0.264 
3. 6555 0.278 
0.9583 0.120 
0.4327 0.033 
0.2503 0.018 


8.3247 0.375 
8. 1696 0.276 
3. 6035 0.251 
0.9077 0.1 12 
0.45 11 0.040 
0.2646 0.018 


8. 3400 0.350 
4.4966 0.403 
1.7251 0.142 
1.28280.074 
0.8968 0.069 
1.08200.275* 


8.3529 0.381 
4.39160.426 
1.6723 0.1 04 
1.2200 0.078 
0.8867 0.068 
1.1390 0.293* 


8.3400 0.350 
7. 4108 0.356 
6.4428 0.390 
7.14260.332 
6.5205 0.375 
7.0346 0.419 


8.3529 0.381 
7. 3072 0.344 
6.7258 0.421 
7.3094 0.348 
6.4219 0.424 
6.9750 0.426 



* Mean from 18 females only. 



76 LEO E. LACHANCE AND SARAH B. BRUNS 

4. Cytology of the irradiated ovary 

To investigate the cytopathological changes related to the inhibition of ovarian 
growth after irradiation, a further series of experiments was conducted. Females 
were irradiated with 2000 or 4000 r, either as 5-day-old pupae or as newly emerged 
adults 0-4 hours old. The ovaries were dissected from the females at 2-i, 4J, and 5^ 
days after emergence and the ovarioles teased apart, fixed, and stained. The slides 
were made permanent and were later examined microscopically for the degree of 
development, presence or absence of nurse cells, presence or absence of second egg 
chambers, and staining differences in the nuclear components, as compared with a 
suitable set of controls. The slides were coded before examination to prevent bias 
in recorded observation. The results are presented in Table III. 

Occasionally in the dissecting and staining procedure, some portion of the 
ovariole was torn or lost. This condition occurred most frequently in the area 
between the first and second egg chambers, which are connected by a very thin 
sheath. Therefore, at times only the first egg chamber could be observed for cyto- 
logical changes after a treatment, and the condition of the rest of the ovariole 
remained unknown. For this reason, there are more first egg chambers recorded 
in Table III than second egg chambers or germaria. In treatment group A, for 
example, 150 ovarioles were examined, but of these only 114 included the second 
egg chamber, and 103 contained the complete ovariole including the germarium. In 
no instance was any component scored as absent if any possibility existed that it had 
been lost in dissection ; when second egg chambers are recorded as absent, both 
the first egg chamber and the germarium were present within the ovariole sheath, 
and the second egg chamber had clearly failed to develop (Fig. 6B). 

Altogether, 1406 ovarioles were examined. The amount of information 
originally collected has been summarized for presentation in Table III. In order 
to organize the data, it was necessary to separate the appearance of ovarioles and 
their contents into categories as follows : 

Reduced normal indicates that the entire structure, including nurse cells, 
appeared normal in shape and degree of staining, but was distinctly smaller in size 
than controls of similar age. This term for the first egg chamber indicates that 
the nurse cells had begun to migrate toward one end of the follicle and that 15 nurse 
cells were present in nearly all instances. For the second egg chamber, all the 
structures were present but reduced in size. Undeveloped refers to the first egg 
chambers and indicates that the chambers were very small but usually contained 
15 nurse cells. Normal development had been retarded so that, in comparison with 
controls of similar age, the ovarioles appeared much younger as well as smaller. 
The nurse cells in the chamber were scattered throughout the entire structure rather 
than tending to gather at the polar end (Fig. 6C). Malformed also refers to first 
egg chambers and is used to describe ovarioles from the older groups (4^ and 5^ 
days old when dissected) that were so badly stunted in size and development that 
they could not even be considered ''undeveloped" (Fig. 7C). Many egg chambers 
were misshapen and contained pycnotic or degenerate nurse cells; the chambers 
were extremely small for their age and showed no signs of development. 

Absent in reference to the second egg chamber indicates a complete lack of a 
recognizable object between the first egg chamber and the germarium, and a vacant 
space in the ovariole sheath (Fig. 6B). Atrophied means that a second egg 



OOGENESIS AND RADIOSENSITIVITY 



77 



g 
- 



a* 
ll 



1 

* 



o 
w 



M a 

~ 1 

^H 

n 



1 



a 9 



^> 

<a 



-8 



i s 



si 



-si 

" 



txa 





0) 

M 

1 


8S 






r^. o\ \o 

(^ 1 ^-H 

CN CN 


>o 
t^. 





-h 


o 
o 

^-H 


p . Os 

CN CN l"~ 

PO i O 


r~ 
t-I 

CN 


arium 


<U 

fin 
Q 


6 
Z 






T-H t^ OO 


o 


SO 


SO 


f*5 00 -H 

<r> < 


CN 

to 


E 

IH 

01 

o 


T) rt 

si 

3 n 


^ 


O 

O 
O 




o 


1 1 


ro '-' * 

OS OO OO 
Os t^ J^ 


10 

c^i 

Os 




O o 



^ 




d 

Os 


O Os ~ 

00 t^ CN 

sO 00 -^ 


ro 
CN 

r^ 




in 


o 
2 


<~r> vo 
o 

T ( 


O 
O 

V ( 


<* 1C Os 
-t 1 CN CN 

^-H 


OO 
Os 


10 Os 

^ 


tf 

to 


O OO OO 

t^. to 


sO 
PC 

** 




o 

.U 

o, 









O 

* 

T ( 


o 

CN 






SO 


"t 

CN 


t_ 
o 


IH 

-4-* 
< 


6 
fc 






\0 


SO 






10 


10 









\O f) 


ro 


i OO ro 


* 


-M t>. 


CN 


Os IO l^- 


CN 


"u 


a 

OJ 
en 

,Q 


&5 


-i CN 
rrj *H 


O 

<~o 


rj< O CN 

T t ^ i 


^-H 


Os 10 

* 00 


Os 

to 


O OO Tf 
"* Os 


f> 
ro 


bg 
o 

13 


< 


6 
fc 


\O 1-1 
ro 


r~- 

ro 


CN ** -^ 
CN 


(^- 

CN 


l^ OO 
CN -H 


10 

TJH 


IO t~ 00 

^ ^H 


O 
I--. 


O 

o 






* "0 


O 


OS CN 1^ 


O 


ON ro 


OO 


1 -rfi ro 


to 


w 


ll 

3 E 


fe< 


OO l~" 

vO 00 


Os 

o 


LO OS r<5 

OO OO 00 


SO 
00 


O -f 5 

IO T-l 




* 


Os IO IO 

to oo 


^f 

sO 




"8 

(* c 


6 
2 


OO 1^- 

t^. 


LO 

00 


** i-O vO 
ro ro rO 
^ i 


<n 
o 

CN 


00 ro 
CN 


^^ 

ro 


to o '-i 

sO l~- 


so 
t*3 

T-H 




o 

V 
kl 


6S 






O *^ 
t--I OO 

to Os 


t*5 

rO 






O fO 

d <rJ 
o 


Os' 
<~O 


I* 


IM 

"to 




6 
2 






10 10 

Tf - 


O 
CN 

T-H 






SO CN 

~f 


00 
* 

v t 


J2 

E 

03 
A 
O 


o 

& 

_o 
u 


fc 


O "0 

C5 CN 

^ 


*-H 




* * 

10 " 


p 

CN 

ro 




d o 



* 1 ^ 1 


p 

d 
o 


O ro 

O ro 
O Os 


ro 

d 

sO 


M 
bo 

CJ 
4J 

2 


> 

1 

p 


6 
2 


O 10 
u-, ^ 

^H 


IO 

\o 

^H 


t^ Os 

o 
^ i 


SO 
* ( 

i i 


CN sO 

sO 
* i 


00 

SO 


CN SO 
f-. IO 


00 
CN 
CN 


E 


ll 

3 E 


* 


1C 

~^ 

OO 


OS 
00 

"0 


\o o o 

OO -H rt 

't 1 <n 


O 

IO 

fO 






ro 
fO 


>O 

d 




a 
t> g 

c* a 


6 



ID 

O 

*-H 


IO 

o 

* f 


IO 
O CN 


t^ 

CN 






CN 


CN 




Number 
observed 




O O 
IO <^I 



!>, 

CN 

"c3 

4-1 

O 

H 


00 Os \O 
O t t 
CN 


*r> 

SO 
fO 

"cd 

4-1 

O 

H 


CN SO 

O so 


00 
sO 

< < 

"3 

-4" 

o 

H 


CN sQ O 
J- f so 


00 
I 
PO 

"t3 

4-> 

^2 




<U I i 
MM* 

< : 5 | 


3^ 
' & 

*3 


rtlN |H|N 

CN -* 




rH|W ^H|N FH[C 

CN ^ IO 




Hw ^IN 
CN Tt 1 




"<|M ^IM ^IC* 

CN Tj< IO 




Treatment 


J* 

< 4 

I- 




<U V 

rt rt 

a a 

3 3 

a a 
x >> 

03 a! 

a -a 

IO IO 




! f] tfl tfl 
-t-> -i-t 4- 

"3 "3 "3 
o-o-o 

a! a3 cd 
u, u i_ 
333 
O O O 
J= -C J= 
i i i 
"*-*"* 
1 1 1 
O O O 




01 <L) 

rt rt 

a a 

3 3 

a a 
>> >> 

rt o3 

o -a 

IO to 




(/) (A CA 



"3 "3 "3 
o -o -c 

rt 03 rt 

1- L. i- 

333 
O O 

JS -C J 

4 1 -4< 4 1 
i i i 
o o o 






_ 

s s 




O 

O 
CN CN 




000 

o o 
o o 

CN CN CN 






O 

* -f 




o o o 

O O 
O 
* Tf rJH 






a 

3 

o 

O 




< 03 




U Q U 




fcO 




E - .-, 





LEO E. LACHANCE AND SARAH B. BRUNS 

chamber was present but small and misshapen, and frequently with no nurse cells 
or very poorly defined ones. 

A reduced normal condition for the germarium means the same as for the first 
and second egg chambers : The structure appeared normal but was much smaller 
in size than controls of the same age. A degenerate germarium indicates that 
there was evidence of the presence of a germarium but that it was so shrunken 
in appearance that it could not be considered merely reduced in size (Fig. 7B). 

Controls for these experiments were untreated females dissected at 2^, 44, and 
5J days after emergence. The appearance of normal ovarioles at various ages has 
been described in section 2, and photographs are given in Figures 6A and 7A. 

Some cytological observations from these experiments could not be readily 
categorized for presentation in Table III, and are discussed in the following 
paragraphs. These comments are intended to supplement the data in Table III. 
The data in the table were based on comparisons of ovarioles from the control 
groups with those from treatment groups. No abnormalities were found in the 
controls, but in none of the treated females was completely normal development of 
an egg chamber ever observed, even at low, non-sterilizing doses of radiation. All 
ovarioles from treated females were distinctly retarded in growth even if nothing 
else appeared abnormal. 

Five-day-old pupae treated u'ith 2000 r (treatment groups A and B). Of 150 
ovarioles dissected from females 2\ days after emergence (group A), all could be 
classified as undeveloped, and 10 of these were very small with tiny nurse cells. 
This pattern was interpreted as retardation of growth rather than cessation, for 
when ovarioles from similarly treated females 4^ days old were examined (group 
B), 87.5% had developed to a stage that would normally be found in control 
females two or three days old that is, the nurse cells were very small but still 
evident and occupying half the volume of the first egg chamber. Of the 270 first 
egg chambers examined in the two groups, all had 15 nurse cells that were reduced 
in size. Undeveloped nurse cells could result in a much slower rate of vitellogenesis 
and thus account for the small or undeveloped egg chambers. The observations on 
second egg chambers indicated that these also formed at a much later time than 
in normal ovarioles. When ovarioles from 2i-day-old females were examined, 
31.6% had failed to develop second egg chambers; of the ovarioles from 4 \- day-old 
females, 87.5% had a second egg chamber but it was small for the age group. 
Apparently the process of ovogenesis in the contents of the ovarioles was retarded 
at least two days by this treatment. The most obvious trend was a slowing down 
of all processes involved in ovogenesis, but not a complete cessation of development. 

Adults 0^ hours old treated with 2000 r (treatment groups C, D, and ). 
Of the 208 ovarioles dissected at 2\ days (group C), all had 15 nurse cells that 
were normal in appearance but reduced in size. In the 79 ovarioles dissected from 
44-day-old females (group D), 45 had malformed ova; 34 had egg chambers that 
were reduced in size or undeveloped, and of these 26 had 15 nurse cells that were 
reduced in size, 4 had less than 15 nurse cells, and 4 had pycnotic or degenerate 
nurse cells numbering 14 or less. In the 76 ovarioles dissected from females at 
5^ days (group E), 98.7% of the first egg chambers examined were malformed, 
and most of these lacked nurse cells or contained shrunken, degenerate nurse cells. 



OOGENESIS AND RADIOSENSITIVITY 79 

In these three treatment groups, 236 ovarioles were complete enough to permit 
observation of second egg chambers, but in only 203 of these were the second egg 
chambers developed. In group D, of 37 complete ovarioles examined, 4 were 
lacking a second egg chamber, 25 had formed a second chamber in which there 
were 15 small nurse cells, 7 had formed a chamber with definitely abnormal nurse 
cells, and one had a chamber with less than 15 nurse cells. Most germaria observed 
in the complete ovarioles were normal in appearance but reduced in size when 
compared with controls of similar age; 16 were completely degenerate and mis- 
shapen. 

In general, the damage to the oocytes was greater after treatment of females 0-4 
hours old with 2000 r than after a like treatment of 5-day-old pupae (see Table II). 
Instead of merely slowing down the process of ovogenesis and allowing the delayed 
production of some normal eggs, treatment of newly emerged females seemed to 
result in a preponderance of malformed eggs. Although the second egg chambers 
were usually developed to some extent (86%), their growth was reduced con- 
siderably and was occasionally followed by atrophy. In addition, degeneration of 
germaria was observed ; this condition did not occur after treatment of 5-day-old 
pupae. 

Five-day-old pupae treated ivitli 4000 r (treatment groups F and G). The 
first egg chambers that developed following treatment with 4000 r grew only very 
slightly and then remained almost stationary in size. Of the 168 first egg chambers 
examined, none had progressed past the "undeveloped" stage. Of 66 first egg 
chambers dissected at 4-J days, 47 contained 15 nurse cells of reduced size, 17 con- 
tained from 2 to 14 small nurse cells of nearly normal appearance, and two contained 
pycnotic or degenerate nurse cells. 

In these two treatment groups, 76 complete ovarioles were observed, but in only 
31 of these were second egg chambers developed. Of the 55 complete ovarioles 
dissected at 2\ days, second egg chambers were absent in 27 ovarioles ; second 
chambers had formed in 28 ovarioles but were extremely reduced in size, and in 
only 8 of these was it possible to observe 15 nurse cells. Of the 21 complete 
ovarioles dissected at 4-i days, 18 did not form second egg chambers, and of the 
remaining three that did, 15 nurse cells were observed in only one and these were 
reduced in size. 

The general trend of ovarian development after treatment of 5-day-old pupae 
with 4000 r was a retardation of growth in the first egg chamber. There were no 
instances of the first egg chamber progressing to the point of formation of an ovum 
without nurse cells ; growth was apparently arrested at an early stage of develop- 
ment. Second egg chambers had failed to develop in 60% of the complete ovarioles 
studied, and both reduction in growth rate and atrophy were observed in the 
germaria. 

Adults 0-4 hours old treated with 4000 r (treatment groups H, I, and /). Of 
the 378 first egg chambers observed, 228 were undeveloped and 148 had developed 
into malformed and abnormal oocytes (Fig. 7C). Of 172 first egg chambers from 
females dissected at 2\ days (group H), all appeared undeveloped, with 15 nurse 
cells of reduced size. Of 146 first egg chambers from 4 '-day-old females (group I), 
all were both undeveloped and malformed ; 58 of the follicles still had 1 5 very small 
nurse cells, but in the remaining 88 the nurse cells had begun to degenerate and 



so LEO E. LACHANCE AND SARAH B. BRUNS 

were difficult to count precisely (Fig. 7C). Of the 60 ovarioles from 5^-day-old 
females (group J), only two could be classified as normal in appearance but dis- 
tinctly retarded in growth ; 56 were still undeveloped, and two malformed. Of the 
first egg chambers from these 60 ovarioles, 35 contained approximately 15 nurse 
cells of reduced size and 25 contained pycnotic or degenerate nurse cells. 

Of 211 complete ovarioles from these three groups, the second egg chamber 
had failed to develop in 70, in 5 it was atrophied, and in 136 it was very much 
reduced in size. There was a relatively high incidence (28%) of germaria being 
found in a degenerate state or absent altogether; in all other instances in which 
the germarium was observed, its size was reduced. 

Previous work has shown that a treatment of 4000 r given to females 0-4 
hours old results in complete sterility (LaChance and Leverich, 1962), whereas 
a similar treatment of 5-day-old pupae results in considerable damage to the repro- 
ductive tissues but does not fully sterilize the females (Bushland and Hopkins, 
1953). 

GENERAL DISCUSSION 

It is tempting to compare the mode of action of the various agents that are 
known to cause sterility. Since Cochliomyia hominivorax is currently being tested 
with a number of chemosterilants, it is hoped that the present observations on the 
pattern of radiation-induced sterility in the female of this species will serve as a 
basis for future comparisons against the effects of chemical sterilization. 

The present studies demonstrate that the effect of irradiation on the reproduc- 
tive capability of female Cochliomyia is largely dependent on the stage of develop- 
ment of the ovarioles at the time the radiation treatment is administered. The 
time during which the egg chambers contain nurse cells undergoing endomitotic 
replications of chromosomal material (adults 0-4 hours old) was the most radio- 
sensitive stage encountered in these studies. Irradiation during this period is much 
more likely to be followed by infecundity than when an equivalent dose of radiation 
is delivered before endomitosis begins (5-day-old pupae) or after it has been 
completed (24-hour-old adults) (see Table II). 

The present studies corroborate the observations of Grosch and Sullivan (1954) 
that the period of endomitosis is especially vulnerable to damage by irradiation and 
that, following treatment at this stage, growth is hampered. There are several 
possible reasons why a group of nurse cells would present an especially vulnerable 
target during the endomitotic process. Although it is true that the polyploid nurse 
cells in very young females present a multiplicity of chromosomal targets as com- 
pared with the diploid oogonial cells in 5-day-old pupae, it is not likely that greater 
radiosensitivity is associated merely with the greater content of DNA in the nurse 
cells. The nurse cells in females 24 hours old certainly contain as much, or more, 
DNA as those of younger females, yet 24-hour-old females are so resistant to 
radiation treatments that even doses of 8000 r do not seriously hamper egg produc- 
tion. Rather, we favor the idea that the failure of the treated females to produce 
mature ova reflects an inability of the nurse cells to support normal vitellogenesis. If 
a radiation treatment is given before the endomitotic replication of nurse cell chromo- 
somal material is completed, the process is likely to be arrested, and the result is the 
formation of nurse cells without the elevated complement of chromosomal material 



OOGENESIS AND RADIOSENSITIVITY 81 

in the nuclei. King and Sang (1959) have suggested that vitellogenesis cannot 
proceed to completion without this elevated chromosomal complement. One of the 
most common features in the present cytological studies of irradiated ovarioles was 
nurse cells that were fairly normal in appearance but very much reduced in size. 

The formation of immature egg chambers with abnormally small nurse cells 
could lead to such changes in the process of oogenesis as a slowing down of 
vitellogenesis or, often, a general cessation of growth at the point at which vitello- 
genesis would normally be most active and contribute most to further increases in 
the size of the oocyte. Occasionally the nurse cells degenerated before growth of 
the oocyte was complete, which resulted in the production of very few eggs or 
malformed eggs that did not hatch. 

On the other hand, there remains the possibility that the oocyte nucleus changes 
in radiosensitivity during the various stages of oogenesis. Damage to the oocyte 
nucleus could conceivably affect the process of vitellogenesis by removing in some 
manner the stimulus for the nurse cells to produce yolk. There is some evidence 
that this factor operates in Drosophila: In studies of ovarian tumors in this species. 
King et al. (1961) observed that nurse cells will form without the presence of an 
oocyte, but that yolk is not produced without an oocyte nucleus in the chamber. 

Ovarian tumors similar to those observed by King (1957) were not found in 
the present experiments with female Cochliomyia. This failure to observe ovarian 
tumors can be attributed to the fact that such tumors are relatively rare (in the 
Drosophila experiments only 13 tumorous egg chambers were observed per 10,000 
cells examined when females were treated with 4000 r). More important, the 
two species differ in the pattern of egg production. Ovarian tumors are of clonal 
origin and arise from oogonial cells that mutate before the formation of a 16-celI 
cyst. Since the egg chambers examined in the present cytological studies were 
derived from cells that had probably already undergone the required number of 
somatic divisions to form a cyst, it was not expected that this type of tumor would 
be found. 

The irradiation of developmental stages in Drosophila has resulted in a reduced 
number of ovarioles, but in Habrobracon juglandis (Ashmead) (= Bracon hebetor 
Say), the normal number of ovarioles always forms after treatment of various, 
growth stages (larvae and pupae) with a number of different radiation doses 
(Erdman, 1961). In this respect, Cochliomyia resembles Habrobracon: The 
present studies clearly indicate that the reduction of growth in the ovary is not due 
to a reduction in the number of ovarioles comprising the ovary, but rather to> 
morphological changes in the ovarioles, which persist in normal numbers. 

Failure of mature ova to form is most often associated with a failure of the 15 
nurse cells to function normally. However, clear instances in which egg chambers 
had formed with less than 15 nurse cells were found after irradiation of both 5-day- 
old pupae and adults 0-4 hours old ; in all such instances growth of the egg chamber 
was seriously hampered. In contrast, King et al. (1961) have shown that in 
Drosophila, vitellogenesis in oocytes can proceed with as few as 10 or as many as 30 
nurse cell nuclei. It is presumed, however, that for a reduced number of nurse 
cells to support vitellogenesis, the nurse cells must contain at least the normal 
polyploid number of chromosomes. 



LEO E. LACHANCE AND SARAH B. BRUNS 

SUMMARY 

1. In normal Cochliomyia hominivorax females, gross ovarian growth, cor- 
related with the age of the adult from emergence to sexual maturity, was measured. 
Studies showed that the size of the ovary doubles between the first and second day 
of adult life, increases more than 5-fold between the second and third day, and 
exhibits a total increase of approximately 60-fold from emergence to sexual 
maturity. 

2. A cytological study of the ovarioles was conducted to determine the sequence 
of events in normal oogenesis. The cytology of the reproductive system, from 5- 
day-old pupae to sexually mature females, is described. 

3. The effects of gamma radiation on gross ovarian growth indicated that newly 
emerged females are more radiosensitive than 5-day-old pupae, and that irradiation 
of 2-day-old females has little effect on subsequent ovarian growth. 

4. The cytopathology of the irradiated ovary was studied after similar doses of 
radiation were delivered to various developmental stages. The general sequence of 
events after treatment was as follows : After low doses, growth is slower than 
normal but not completely arrested. If treatment is given to 5-day-old pupae, 
grossly malformed oocytes are not often encountered, but second egg chambers fre- 
quently do not form. When females 0-4 hours old are irradiated, growth in the 
first egg chamber is delayed considerably, and is often followed by complete de- 
generation of the first egg follicle or the formation of grossly malformed oocytes. 
In these studies, the most radiosensitive stage encountered was that period during 
which the egg chambers contain nurse cells undergoing endomitotic replications of 
chromosomal material. 

LITERATURE CITED 

BAUMHOVER, A. H., A. J. GRAHAM, B. A. BITTER, D. E. HOPKINS, W. D. NEW, F. H. DUDLEY 

AND R. C. BUSHLAND, 1955. Screw-worm control through release of sterilized flies. 

/. Econ. Ent., 48: 462-466. 
BIER, K., 1960. Der Karyotyp von Calliphora erythrocephala Meigen unter besonderer Bertick- 

sichtigung der Nahrzellkernchromosomen im gebundelten und gepaarten Zustand. 

Chromosoma, 11: 335-364. 

BUSHLAND, R. C., 1960. Screw-worm research and eradication. Advances in Vet. Sci., 6: 1-18. 
BUSHLAND, R. C., AND D. E. HOPKINS, 1953. Sterilization of screw-worm flies with X-rays and 

gamma rays. /. Econ. Ent., 46 : 648-656. 
ERDMAN, H. E., 1961. Analyses of the differential radiosensitivity of developing reproductive 

tissues in Habrobracon juglandis (Ashmead) to ionizing radiations. Int. J. Rad. Biol., 

3: 183-204. 
FAWCETT, D. W., S. ITO AND D. B. SLAUTTERBACK, 1959. The occurrence of intercellular bridges 

in groups of cells exhibiting synchronous differentiation. /. Biophys & Biochem. 

Cytol, 5: 453-460. 
GROSCH, D. S., AND R. L. SULLIVAN, 1954. The quantitative aspects of permanent and temporary 

sterility induced in female Habrobracon by x-rays and beta radiation. Rad. Research, 

1: 294-320. 
HIRSCHLER, J., 1955. On the cooperation of fusomes in the development of egg-nurse cell 

complexes in the animal ovary. La Cellule, 57: 67-87. 
JACOB, J., AND J. L. SIRLIN, 1959. Cell function in the ovary of Drosophila melanogaster. I. 

DNA classes in the nurse cell as determined by autoradiography. Chromosoma, 10: 

210-228. 

JEFFERSON, M. E., 1960. Irradiated males eliminate screw-worm flies. Nucleonics, 18: 74-76. 
KING, R. C., 1957. Oogenesis in adult Drosophila melanogaster. III. Radiation-induced ovarian 

tumors. Grozvth, 21: 129-135. 



OOGENESIS AND RADIOSENSITIVITY 83 

KING, R. C., 1960. Oogenesis in adult Drosophila melanogaster. IX. Studies on the cyto- 
chemistry and ultrastructure of developing oocytes. Grotvth, 24: 265-323. 

KING, R. C., AND J. H. SANG, 1959. Oogenesis in adult Drosophila melanogaster. VIII. The 
role of folic acid in oogenesis. Growth, 23: 37-53. 

KING, R. C., E. A. KOCH AND G. A. CASSENS, 1961. The effect of temperature upon ovarian 
tumors of the Fes mutant of Drosophila melanogaster. Growth, 25 : 45-65. 

LACHANCE, L. E., AND A. P. LEVERICH, 1962. Radiosensitivity of developing reproductive cells 
in female Cochliotnyia hominivorax. Genetics, 47 : 721-735. 

LORZ, A. P., 1947. Supernumerary chromonemal reproductions : Polytene chromosomes, endo- 
mitosis, multiple chromosome complexes, polysomaty. Botan. Reviews, 13 : 597-624. 

PAINTER, T. S., 1959. Some values of endomitosis. In: Biological Contributions. M. R. 
Wheeler (ed.), Univ. of Texas Publ. #5914, University Press, Austin, pp. 235-240. 

PAINTER, T. S., AND E. C. REINDORP, 1939. Endomitosis in the nurse cells of the ovary of 
Drosophila melanogaster. Chromosoma, 1: 276-283. 

STALKER, H. D., 1954. Banded polytene chromosomes in the ovarian nurse cells of adult Diptera. 
/. Heredity, 45: 259-264. 



PHOTOPERIODIC TERMINATION OF DIAPAUSE IN AN INSECT 1 

D. G. R. McLEOD 2 AND STANLEY D. BECK 

Department of Entomology, University of Wisconsin, Madison 6, Wisconsin 

Since the work of Kogure (1933) on the effect of photoperiod on diapause in 
the commercial silkworm, diapause in many insect species has been found to be 
photoperiodically induced (see reviews by Lees, 1955; de Wilde, 1962). The 
European corn borer, Ostrinia nubilalis, displays a photoperiodically induced facul- 
tative diapause in the final (fifth) larval instar (Mutchmor and Beckel, 1958, 
1959; Beck and Hanec, 1960). 

Diapause is defined as a state of arrested development in which the arrest is 
enforced by a physiological mechanism (Beck and Hanec, 1960). Diapause is, 
therefore, distinguishable from quiescence or dormancy that is enforced by un- 
favorable environmental conditions. Under natural conditions, diapause is eventu- 
ally terminated and morphogenesis resumed. The physiological processes involved 
in the termination of diapause constitute developmental changes on a biochemical 
level, and have been termed "diapause development" by Andrewartha (1952). 
Andrewartha defined diapause development as (1952, p. 53) "the physiological 
development, or physiogenesis, which goes on during the diapause stage in prepara- 
tion for the active resumption of morphogenesis." It is, therefore, the process of 
reversing (or replacing) the physiological mechanism enforcing the diapause state. 

Experimental work on diapause development has dealt mainly with the low- 
temperature treatments necessary to terminate diapause; the reviews of Andre- 
wartha (1952) and Lees (1955) discuss many examples of diapause development 
in eggs, larvae, pupae, and adults. In a few instances, termination of diapause has 
been found to be photoperiodically induced without a previous exposure of the 
insects to low temperatures (Baker, 1935; Paris and Jenner, 1959; Shakhbazov, 
1961). According to the definitions employed above, photoperiodic termination 
of diapause must also involve diapause development. 

The intensity of diapause, as measured by the length of time required to com- 
plete diapause development, varies widely among the species studied; a few days 
are required for Lo.vostege sticticalis (Pepper, 1937), but several months are 
needed in the case of Melanoplus bivittatus (Church and Salt, 1952). The in- 
tensity of diapause may also vary among individuals of the same species, depending 
on how long they have been exposed to diapause-inducing conditions. De Wilde 
et al. (1959) reported that diapause in the Colorado potato beetle, Leptinotarsa 
decemlincata Say, was photoperiodically reversible shortly after the adults displayed 
diapause behavior, but not a few days later. Hogan (1962) found that embryonic 

1 Approved for publication by the director of the Wisconsin Agricultural Experiment Station. 
This study was supported in part by a research grant (RG-7557) from the National Institutes 
of Health. 

2 Present Address : Canada Department of Agriculture, Entomology Laboratory, Chatham, 
Ontario. 

84 



INSECT DIAPAUSE TERMINATION 85 

diapause in the cricket Acheta commodus was more intense after 14 days of 
incubation at 23 C. than after only 7 days at that temperature. 

Babcock (1924) reported that larval diapause in the European corn borer 
could be terminated only after the larvae had been subjected to about 6 weeks of 
freezing or near-freezing temperatures. This finding has been generally accepted, 
and in the absence of any photoperiodic treatment of the diapause larvae, is 
verifiable. In a recent study, Beck and Apple (1961) reported that diapause could 
be revoked in laboratory-reared borers by subjecting them to long-day photoperiods 
shortly after they had entered the diapause state. A low-temperature exposure 
was not required for such diapause termination, and the authors expressed doubt 
as to whether or not the larvae had been fully in diapause. They referred to such 
easily broken diapause as an "incipient" rather than "true" diapause. 

The study here presented was undertaken in an effort to determine whether or 
not diapause in the European corn borer can vary in intensity under different con- 
ditions, and also to elucidate the relationship between photoperiod and the com- 
pletion of diapause development. 

MATERIAL AND METHODS 

The European corn borers used in this study were from a restricted natural 
population occurring near Madison, Wisconsin. The use of a defined population 
was necessary because of the demonstration of significant differences in photo- 
periodic responses among different geographical populations of this species (Beck 
and Apple, 1961). Overwintering borers were collected from the field in the fall 
of the year, and were stored at 5 C. As needed, groups of stored borers were 
incubated at 30 C. for pupation and emergence. The progeny of these insects 
were used in the experiments described below, except where field borers are 
indicated. The times of collection and storage conditions for field borers are 
indicated in the appropriate sections. 

The laboratory borers were reared aseptically on purified diets according to the 
rearing techniques described by Beck and Smissman (1960). All experiments 
were run at 30 C. with the exception of those treatments involving a temperature 
cycle or storage at 5 C. 

The experiments were carried out in B.O.D. constant temperature incubators 
that had been modified to incorporate a thermistor temperature control system 
(Thermistemp Temperature Control Model 71, Yellow Springs Instrument Com- 
pany, Yellow Springs, Ohio). Control of photoperiod was effected through the use 
of 7-day cycle programmers wired to two 14-watt fluorescent lights installed in the 
incubator. In experiments involving temperature changes, two methods were 
used. Symmetrical temperature cycles were obtained by using a clock motor to 
drive the thermistor temperature control unit through a prescribed cycle. The 
temperature reached a maximum of 31 C. and 12 hours later reached a minimum 
of 21 C. A temperature cycle with abrupt changes was obtained by having the 
thermistor temperature controller switched from 32.5 C. to 12 C. by a 24-hour 
programmer. This apparatus was set to give a temperature cycle with 16 hours at 
32.5 C. and 8 hours at 12 C. The change from the maximum temperature to 
the minimum temperature took 1-J hours, while the change from the minimum to 



86 



D. G. R. McLEOD AND STANLEY D. BECK 



the maximum temperature took one hour. The performance of the temperature- 
controlling apparatus was verified by a recording thermograph. 

Throughout this paper the term short-day refers to a photoperiod consisting of 
13 hours of photophase and 11 hours of scotophase, and the term long-day refers to 
a photoperiod with a 16! -hour photophase and a 7^-hour scotophase. These experi- 
mental conditions were employed because previous work had shown that over 90% 
incidence of diapause was induced in the Madison population of the European corn 
borer when they were grown under a 12-13-hour photophase in a 24-hour photo- 
period. Either more or less light induced less diapause. There was no appreciable 

l.4r 



1.2- 



(D 



cr 

LJ 

a 



1.0 



.8 



.6 



NON DIAPAUSE 




DIAPAUSE 



_L 



10 



14 18 22 26 

LARVAL AGE IN DAYS 



30 



34 



FIGURE 1. Rate of oxygen consumption by diapause and nondiapause 
larvae of the European corn borer. 

diapause in response to a long day, continuous light, or continuous dark (Beck and 
Hanec, 1960). 

Diapause in the individual borer was determined by a negative criterion fail- 
ure to pupate. With such a criterion, there had to be an arbitrary point of time 
selected, after which the borers were considered to be in diapause. This point 
was reached experimentally when the control borers, reared in the dark, had 
finished pupating and the pupation curve for the experimental population had 
leveled off (see Figure 2). Diapause borers were obtained for experimentation 
by rearing them in a short day for 21 days after eclosion. At this time they were 
placed in clean vials on wet paper strips. Experiments on breaking diapause were 
terminated when all the borers in the sample had either pupated or died. 



INSECT DIAPAUSE TERMINATION 



87 



The oxygen consumption of laboratory-reared borers was carried out using 
standard manometric techniques (Umbreit et al., 1957). Borer larvae were con- 
fined in small wire cages in each Warburg flask. This prevented them from 
crawling into the center well and, because the insect is thigmotactic, also tended to 
keep them relatively inactive. The oxygen consumption of each larva was measured 
for one hour in 10-minute increments. The p\. O 2 consumed per hour was divided 
by the live weight of the insect to give /,!. O 2 /mg./hr. 

The analyses of variance were calculated by the method of Steel and Torrie 
(1960) for groups with unequal replication. Duncan's New Multiple Range test 



9 HR PHOTOPHASE 



13 HR PHOTOPHASE 




10 



20 



30 40 50 

LARVAL AGE IN DAYS 



60 



70 



80 



FIGURE 2. Pupation curves for European corn borer populations reared under different 
photoperiods. The arrow indicates the time at which all diapause larvae were exposed to long- 
day photoperiods (16.5L/7.5D). 

was used with the approximation of Kramer (cited in Steel and Torrie, 1960) for 
testing means based on unequal replication. 

RESULTS AND DISCUSSION 

The diapause stage in most insects is characterized by a low level of oxygen con- 
sumption (Heller, 1926; Boell, 1935; Schneiderman and Williams, 1953). Beck 
and Hanec (1960) found that the respiration of borers collected from the field also 
dropped, but little was known about the length of time necessary to reach this low 
level of oxygen consumption as the borers entered diapause. The "incipient" 
diapause reported by Beck and Apple (1961) may have reflected an incomplete 
suppression of respiration. Once a stable low level of oxygen consumption was 
reached, the "true" diapause stage would have been attained. 

Figure 1 shows that oxygen consumption declined after the moult from the 
fourth to the fifth instar in both diapause and nondiapause borers. The non- 



D. G. R. McLEOD AND STANLEY D. BECK 

diapause respiration increased at the pupal moult to form the classical U-shaped 
curve. Respiration in the diapause larvae continued to drop, stabilizing at from 
one half to one third the prediapause level. If a stable low level of oxygen con- 
sumption is indicative of diapause, then the diapause stage had been reached at this 
time (22-24 days). If diapause occurred in different intensities after this time, 
measurement of its intensity would require some method other than oxygen con- 
sumption. The average time to pupation when placed in a long day was the cri- 
terion used to determine the intensity of diapause in subsequent experiments, the 
supposition being that the more intense the diapause, the greater the delay 
in pupation. 

The intensity of diapause in larvae of the European corn borer was tested in 
larvae reared under different photoperiods (Fig. 2). Diapause incidence of greater 
than 90% was observed among larvae reared under a photoperiod consisting of a 
13-hour photophase and an 11 -hour scotophase. Intermediate incidence of diapause 
was obtained under a 15-hour photophase and a 9-hour scotophase and also under 
a 9-hour photophase and 15-hour scotophase. The 15-hour photophase is on the 
long-day side of the response maximum, and the 9-hour photophase is on the short- 
day side of the response peak (Beck, 1962). 

TABLE I 

The average time to pupation of diapause borers grown tinder different 
photophases when transferred to a long day 

Photophase Average time to 

(hr./24 hours) pupation 

9 29.9 

13 29.8 

15 13.5* 

* This mean is significantly different from the other means at the 5% level of probability. 

Larvae reared under the 9- and 13-hour photophase treatments reached a maxi- 
mum incidence of pupation at 15-16 days, and the pupation curve leveled off, while 
the 15-hour treatment leveled off more slowly (24 days). The dark controls had 
finished pupating by 22 days. The non-pupators in the three treatments were then 
considered to be in diapause and were placed in a long-day photoperiod. 

When exposed to a long-day photoperiod, the borers reared under the 9- and 
13-hour photophase began to pupate in 17-19 days. At from 65 to 70 days of age, 
all had pupated. The 15-hour photophase treatment, on the other hand, did not 
result in this pattern. Diapause in these larvae did not appear to be as intense as 
in the other two groups, and pupation was nearly finished by the time that pupa- 
tion had started in the other two experimental populations. The average times to 
pupation, as shown in Table I, were also significantly lower. Different intensities 
of diapause occurred under different photoperiods, but long-day exposures termi- 
nated diapause in borers grown under any of the three photoperiods tested. 

There still remained the possibility that longer exposure to a short day would 
induce a more intense diapause, as shown by de Wilde et al. (1959) and Hogan 
(1962). To test this hypothesis, diapause was induced in a large group of borers 
with a short-day photoperiod. At 30 days of age, and at subsequent 20-day in- 
tervals to 90 days, samples were removed and placed in either continuous dark or 



Cont. dark 
(days) 


Long day 
(days) 


78 (2)* 
29 (3) 

17 (3) 


30.7 (23) 

30.1 (12) 
21.7 (12)** 
16.1 (14)** 



INSECT DIAPAUSE TERMINATION 89 

TABLE II 

The average time to pupation of diapause larvae sampled at different ages 
from a population held continuously under a short day 

Average time to pupation 
Larval age 
at transfer 
(days) 

30 
50 
70 
90 

* Numbers in brackets refer to the number of pupae/sample of 30. 
** This mean significantly different from the rest at the 5% level. 

a long-day regime, and observed daily for pupation (Table II). Diapause was 
terminated at any age, and the average time to pupation became shorter with in- 
creasing age. The average time to pupation in the 70- and 90-day samples trans- 
ferred to the long day may have been shortened partly because some of the borers 
were close to pupation before the long-day treatment was begun. 

Diapause in the corn borer does not last indefinitely, even under diapause- 
inducing conditions. The borers that remained under the diapause-inducing short 
day started to pupate at about 70 days of age, and by 140 days 36% had pupated 
and 64% had died. This finding can, perhaps, be explained on the basis that 
diapause development may proceed at 30 C. Diapause development then proceeds 
to completion in about one third of the population by 140 days. This would mean 
that diapause in the borer is much the same as that in Philosamia cynthia (Dani- 
lyevsky, 1949; cited by Lees, 1955). The high end of the temperature range for 
diapause development corresponds with a large part of the temperature range for 
morphogenesis. Although diapause development can occur at 30 C., the high rate 
of mortality and the length of time for pupation would indicate that this is not 
the optimum temperature for diapause development. It is probable that mortality 
occurs because the borers have utilized their fat body reserves before diapause de- 
velopment has been completed. An alternative hypothesis is that a diapause of 
sufficient intensity was not attained under the experimental conditions. 

Experiments were set up to determine if a more intense diapause could be 
induced by varying the temperature and photoperiod to which the insects were 
exposed. Diapause was induced in two groups of borers, and they were placed in 

TABLE III 

The average time to pupation of diapause borers stored in continuous 
dark at 30 C. when transferred to a long day 

Larval age at Days in Average time to pupation 

transfer dark in long days 

(days) (days) 

21 29.3 (24)* 

35 14 29.0 (22) 

49 28 24.2 (25) 

63 42 28.0 (18) 

77 56 27.1 (16) 

* Numbers in brackets refer to the number of pupae/sample of 30. 



90 



D. G. R. McLEOD AND STANLEY D. BECK 



continuous dark at either 30 C. or 5 C. when they were 22 days old. Periodically 
samples were removed and placed in a long day, and the average time to pupation 
determined. Tables III and IV show that neither continuous dark nor cold treat- 
ment induced a more intense diapause. Diapause was terminated by a long day, 
and there was no requirement for chilling. The F value from the analysis of vari- 
ance for the data in Table III was not significant at the 5% level. Table IV shows 
that the long-day treatments yielded significantly different results. The borers 
that were exposed to 5 C. for 42 days took less time to pupate than any of the 
other three groups. But it can be seen from the continuous-dark column (Table 
IV) that diapause development was completed in very few of the borers during 
the exposure to 5 C. 

In the field, corn borers are not exposed to a constant temperature as in the 
incubator but to a fluctuating temperature with a low during the night and a high 
during the day. Beck (1962) found that 96% diapause was induced when borers 
were grown under a short-day photoperiod and a thermoperiod with the cold 
during the dark. If the cold came during the light, only 15% diapause was 
induced. 

TABLE IV 

The average time to pupation of diapause borers stored in continuous dark at 5 C. 
when transferred to a long day or continuous dark at 30 C. 



Larval age at 




Average time to pupation in: 










(5 C ) 










Cont. dark 


Long day 


(days) 




(days) 


(days) 


21 





(0)* 


27.2 (21) 


35 


14 


52.4 (5) 


28.1 (18) 


49 


28 


43.0 (3) 


30.7 (9) 


63 


42 


33.0 (1) 


19.4 (12)** 



* Numbers in brackets refer to the number of pupae/sample of 30. 
** This mean is significantly different from the others at the 5% level of probability. 

Diapause was induced in a short-day and either a constant temperature (26 
C.) or a 21-31 C. symmetrical thermoperiod. The average time required for 
pupation was compared in the two groups by samples taken at two different times 
(26 and 50 days). The F values from the split plot analysis of variance were not 
significantly different at the 5% level for the long-day treatments in Table V. 
Thus, a thermoperiod imposed on a photoperiodic regime did not induce a more 
intense diapause. 

A diapause incidence of about 45% was induced when borers were reared in 
continuous dark and a thermoperiod consisting of 16 hours at 32.5 C. and 8 hours 
at 12 C. with abrupt temperature changes. Borers were reared and held under 
these conditions for 90 days. Periodically samples were removed and placed in 
either continuous dark or long day, and observed for pupation. Table VI shows 
the average time to pupation for these groups. The analysis of variance for the 
long-day treatments gave a non-significant F value at the 5% level, and it is 
concluded that a thermoperiod alone does not induce a more intense diapause. 



INSECT DIAPAUSE TERMINATION 



91 



TABLE V 

The average time to pupation of borers grown under a short day and either (A) 
constant temperature, or (B) 21-31 C. symmetrical thermoperiod when 
transferred to a long day or continuous dark 



Larval age at 




Average time to pupation 










condition 


Cont. dark 


Long day 


(days) 




(days) 


(days) 


26 


A 


108 (1)* 


29.8 (25) 


26 


B 


68 (1) 


28.2 (23) 


50 


A 


41.5 (5) 


28.3 (27) 


50 


B 


59.8 (5) 


33.1 (18) 



* Numbers in brackets refer to the number of pupae/sample of 30. 

The preceding data show that an intense diapause, that is, a diapause that re- 
quired a prolonged period of chilling for termination, could not he induced. Dia- 
pause could always be terminated with a long-day photoperiod. There was no 
requirement for chilling to break diapause. 

Experiments on the response of field-collected borers to a long-day photoperiod 
were begun at the end of August when pupation was completed in the field popula- 
tion. Samples of borers were collected periodically and placed in either a long 
day or in continuous dark. Table VII shows the average time to pupation for 
these groups. The results show that diapause could be broken in the field popula- 
tion by a long day, and that diapause termination did not require chilling. It is 
evident that the diapause induced in laboratory-reared borers was similar in 
intensity to that of the field-collected borers. 

It is evident from the "continuous dark" column (Table VII) that diapause 
development had been completed in the majority of borers by December 13. The 
average time to pupation became shorter after this date, because morphogenesis 
proceeded slowly at the low temperature in the field. After this date, the average 
time to pupation in the long day did not differ significantly from that in the dark. 
Thus, once diapause development had been completed, long-day exposure had no 
further effect. 

The experimental results discussed thus far show that diapause development 

TABLE VI 

The average time to pupation of diapause borers grown in continuous dark and a 

thermoperiod of 16 hours 32.5 C. and 8 hours 12 C. when placed in 

continuous dark or long day 



Larval age at 

transfer 

(days) 

45 
60 

75 
90 



Average time to pupation 



Cont. dark 
(days) 

68.3 (3/15)' 



21.6 (5/15) 



Long day 
(days) 

31.0 (13/15) 

31.7 (18/30) 
26.9 (17/30) 

23.1 (14/25) 



* Numbers in brackets refer to the number of pupae/number in sample. 




92 D. G. R. McLEOD AND STANLEY D. BECK 

can proceed very slowly to completion at 30 C. in continuous dark or under a 
short day. Long-day treatment greatly accelerates the rate of diapause develop- 
ment. The mortality that occurred when borers were held for a long period of 
time at 30 C. indicates that borers may need chilling, not to complete diapause 
development, but rather to decrease the metabolic rate so that there is enough fat 
body reserve to insure morphogenesis once diapause devlopment has been completed. 
Since diapause development can be completed in the corn borer by a long-day 
photoperiod, a question arises as to the number of days of long-day treatment 
necessary. To answer this question, diapause was induced in a group of borers 
by rearing them under a short-day photoperiod for 21 days; they were then 
placed in a long-day incubator. Periodically samples were removed and placed in 
continuous dark and the percentage pupation observed at 66 days of age. The ex- 
periment was terminated at this age because pupation was finished in the group that 

TABLE VII 

The average time to pupation of diapause field-collected borers when placed in 
either a long day or continuous dark 

Average time to pupation 
Date 
collected 

Aug. 22 
Sept. 6 
Sept. 20 
Oct. 4 
Oct. 18 
Nov. 1 
Nov. 15 
Nov. 29** 
Dec. 13 
Dec. 27 
Jan. 10 
Jan. 24 
Feb. 7 
Apr. 24 

* Numbers in brackets refer to the number of pupae/sample size. 
** This collection was heavily infected with a fungus and high mortality occurred. 

remained in the long day, and pupation could occur after 70 days among diapause 
borers held in the dark. 

Figure 3 shows that diapause was terminated in 20% of the population by as 
little as two days of long-day treatment, and with 12-16 days, 100% pupation oc- 
curred. It is evident that diapause development induced by the long-day photo- 
period took up little of the time (20-30 days) necessary to reach the pupal stage; 
the remainder of the time was required for morphogenesis. 

The terms "diapause" and "diapause development" have been used in this dis- 
cussion with no attempt to assign these terms to any physiological mechanism, al- 
though experimentally these processes have been shown to occur. Wigglesworth 
( 1934) was the first to propose that diapause was linked to the hormones controlling 
growth and metamorphosis. Subsequently, Williams (1946, 1947, 1948) showed 
that diapause in the Cecropia silkworm was caused by the failure of the neuro- 



Long day 


Cont. dark 


(days) 


(days) 


24.7 (25/27)* 


71.1 (9/24) 


24.6 (28/30) 


93.5 (4/6) 


25.9 (22/30) 


124 (1/6) 


27.7 (26/30) 


60.0 (2/6) 


24.9 (18/21) 





27.0 (24/30) 


73.6 (7/21) 


27.4 (27/30) 


57.7 (4/21) 


24.0 (3/30) 


29.0 (1/30) 


26.8 (25/36) 


26.6 (17/36) 


21.2 (20/30) 


26.9 (14/30) 


16.7 (21/30) 


20.2 (17/30) 


14.9 (21/30) 


16.6 (24/30) 


15.1 (21/30) 


17.4 (17/30) 


11.8 (22/27) 


11.8 (16/27) 



INSECT DIAPAUSE TERMINATION 



93 



secretory cells of the brain to secrete the activation hormone. The brain regains 
its ability to secrete the activation hormone when chilled for 6 weeks at 5 C. 
The activation hormone stimulates the prothoracic glands to secrete ecdyson, and 
adult differentiation follows. In this case "diapause" refers to the "metabolic 
block" preventing secretion of the activation hormone ; and "diapause development" 
to the process taking place at 5 C. that returns the brain to secretory activity. This 
general scheme has been found to fit the larval diapause in the wheat stem sawfly, 
Cephus cinctus (Church, 1955) and the pupal diapause in the Lime hawk moth, 
Mimas tiliae (Highnam, 1958). 

It has been found by Cloutier et al. (1962) that the brain of the diapause 
borer is able to promote adult development in diapause larvae. One of the hy- 



100 



80 



UJ 

O 



S 60 
Q. 



40 



Q. 



20 



O 



O 







10 



INDUCTION PERIOD IN DAYS 



15 



20 



FIGURE 3. Effect of different long-day exposures on the termination 
of diapause in the European corn borer. 

potheses presented to explain this was that the brain contained the activation hor- 
mone and that its release was blocked in the intact diapause brain. This situation 
differs from that found in the giant silkworm, Hyalophora cecropia; Williams 
(1948) found that up to 8 diapause brains could not promote adult development in 
diapause pupae. Thus, it appears that diapause development in the European 
corn borer facilitates the release of the product, and not its production. 

The sequence of events that appears to take place before pupation can occur is : 
(1) removal of the block preventing release of neurosecretory products; (2) secre- 
tion of the activation hormone in sufficient titer to stimulate the prothoracic glands ; 
and (3) production and secretion of ecdyson in sufficient titer to promote adult 
differentiation. It is well substantiated that once the prothoracic glands have been 
stimulated sufficiently by the activation hormone, adult differentiation can proceed 



D. G. R. McLEOD AND STANLEY D. BECK 

TABLE VI II 

The per cent pupation after the re-induction of diapause 

Pupation when transferred to 



Days in Short day Cont. dark 

long day % % 



7 48 

9 33 53 

11 67 79 

Continuously in long day 100 

without further brain activity (Williams, 1946). This is also true for the corn 
borer. Of 24 pupae that had their brains removed immediately after pupation, the 
22 survivors underwent adult differentiation and emerged as moths. This showed 
that sufficient activation hormone had been liberated prior to the pupal moult to 
stimulate adult differentiation, and that the prothoracic glands could be effective in 
the absence of the brain. If, in terminating diapause with a photoperiodic cycle, 
secretion of the activation hormone takes appreciable time to reach a level high 
enough to stimulate the prothoracic glands, there might be some point after dia- 
pause development has been completed where the block to secretion might be re- 
established by a short day. Thus, it could be possible to re-induce diapause by 
physiologically "removing" the brain. 

To test this hypothesis, diapause borers were placed in a long day to induce 
diapause development. Periodically duplicate samples were removed and placed 
in either continuous dark or short day. After 45 days from the beginning of the 
long-day treatment, the percentage pupation was calculated in all groups. Table 
VIII shows that diapause could be re-induced in all those borers that had only 7 
days of long-day treatment and a small percentage could be made to re-enter dia- 
pause after 9 and 11 days of long-day treatment. This experiment does not take 
into account the time necessary to establish the block to neurosecretion, but at least 
7 days is not time enough to stimulate the prothoracic glands. But from those 
samples placed in continuous dark, it can be seen that 7 days of long-day treat- 
ment was enough to terminate diapause in 48% of the population. It would thus 
appear that secretion of the activation hormone is a part of morphogenesis and not 
diapause development, and hence is not effected by photoperiod. 

In the previous experiment the borers were transferred to a short day and 
remained there until the end of the experiment. As diapause could be reinstated 
in all after 7 days of long-day treatment, this time was used to determine how many 

TABLE IX 

The per cent pupation at 45 days when different lengths of time for diapause 

re-induction were used 

Days in Days in Per cent 

long day short day pupation 

7 46 

7 4 40 

7 8 20 

7 12 6 



INSECT DIAPAUSE TERMINATION ( ^ 

days of short treatment were necessary to re-induce diapause. This experiment 
was set up in the same manner as the previous one except that the borers were 
removed from the short-day incubator at various times and placed in continuous 
dark. The per cent pupation was calculated in all groups 45 days after the transfer 
to the long day (66 days of age). 

Table IX shows that at least 12 days of short-day treatment are necessary to re- 
duce the per cent pupation from 46% to 6%. It appears, then, that considerable 
time is required to re-establish the block to neurosecretion. The short length of 
time necessary for diapause development in a long day compared with the length of 
time necessary to induce diapause is quite in agreement with what is known about 
the original induction of diapause (Beck and Hanec, 1960). 

Borers collected from the field on October 28 and stored at 5 C. in the dark 
for 53 days were subjected to different photoperiods. The average times to pupa- 
tion in Table X were not significantly different at the 5% level. Thus, photoperiod 

TABLE X 

Average time to pupation of field-collected borers under different photoperiods 

Photoperiodic Average time 

condition to pupation 

Continuous dark 18.3 

Continuous light 15.4 

Long day 15.7 

Short day 13.9 

had no effect on these borers. These findings imply that diapause development had 
been completed, and that morphogenic development had advanced beyond the 
point where it could be arrested. 

SUMMARY 

1. The European corn borer, Ostrinia nubilalis, has a faculative diapause in the 
last larval instar. Diapause induced in the laboratory by a short-day photoperiod 
is identical in its intensity to that occurring in the field. 

2. Diapause development occurs at 30 C. under various photoperiodic condi- 
tions but is greatly accelerated by a long day. 

3. Completion of diapause development does not require a period of chilling. 

4. Diapause development is a process that removes a block to secretion of the 
activation hormone but does not include secretion or any of the morphogenic 
events that follow. 

LITERATURE CITED 

ANDREWARTIIA, H. G., 1952. Diapause in relation to the ecology of insects. Biol. Rev., 27: 

50-107. 
BABCOCK, K. W., 1924. Environmental studies of the European corn borer (Pyrausta nubilalis). 

J. Econ. Ent., 17: 120-125. 
BAKER, F. C., 1935. The effect of photoperiodism on resting tree-hole mosquito larvae. 

Canad. Ent., 149-153. 

BECK, S. D., 1962. Photoperiodic induction of diapause in an insect. Biol. Bull., 122: 1-22. 
BECK, S. D., AND J. W. APPLE, 1961. Effect of temperature and photoperiod on voltinism of 

geographical populations of the European corn borer, P \rausta nubilalis (Hbn.). 

/. Econ. Ent., 54: 550-558. 



96 D. G. R. McLEOD AND STANLEY D. BECK 

BECK, S. D., AND W. HANEC, 1960. Diapause in the European corn borer, Pyrausta nubilalis 

(Hbn.). /. Ins. Physiol, 4: 304-318. 
BECK, S. D., AND E. E. SMISSMAN, 1960. The European corn borer, Pyrausta nubilalis (Hbn.), 

and its principal host plant. VIII Laboratory evaluation of host resistance to larval 

growth and survival. Ann. Ent. Soc. Amer., 53: 755-762. 
BOELL, E. J., 1935. Respiratory quotients during embryonic development (Orthoptera). /. Cell. 

Comp. Physiol., 6: 369-385. 
CHURCH, N. S., 1955. Hormones and termination and reinduction of diapause in Cephus cinctus 

Nort. (Hymenoptera : Cephidae). Canad. J. Zool., 33: 339-369. 
CHURCH, N. S., AND R. W. SALT, 1952. Some of the effects of temperature on development and 

diapause in eggs of Melanoplus bivittatus (Say) (Orthoptera: Acrididae). Canad. J. 

Zool., 30: 173-184. 
CLOUTIER, E. J., S. D. BECK, D. G. R. McLEOD AND D. L. SILHACEK, 1962. Neural transplants 

and insect diapause. Nature, 135 : 1222-1224. 
HELLER, J., 1926. Chemisch Untersuchungen tiber die Metamorphose der Insekten. Ill Mit- 

teilung: Uber die 'subitane' und 'latente' Entwicklung. Biochem. Zeitschr., 169: 

208-234. 
HIGH NAM, K. C, 1958. Activity of the brain/corpora cardiaca system during pupal diapause 

'break' in Mimas tiliae (Lepidoptera). Quart. J. Micr. Sci., 99: 73-88. 
HOGAN, T. W., 1962. The absorption and subsequent breakdown of urea by diapausing eggs of 

Acheta commodus (Walk.) (Orthoptera: Gryllidae). Aitst. J. Biol. Sci., 15: 362-370. 
KOGURE, M., 1933. The influence of light and temperature on certain characters of the silkworm, 

Bombyx mori. J. Dept. Agric., Kyushu Univ., 4: 1-93. 
LEES, A. D., 1955. The Physiology of Diapause in Arthropods. Cambridge University Press, 

Cambridge. 
MUTCHMOR, J. A., AND W. E. BECKEL, 1958. Importance of photoperiod and temperature in 

inducing diapause in the European corn borer, Pyrausta nubilalis (Hbn.). Nature, 

181 : 204. 
MUTCHMOR, J. A., AND W. E. BECKEL, 1959. Some factors affecting diapause in the European 

corn borer, Ostrinia nubilalis (Hubn.). Canad. J. Zool., 37: 161-168. 
PARIS, O. H., AND C. E. JENNER, 1959. Photoperiodic control of diapause in the pitcher-plant 

midge, Metriocnemus knabi. Symposium on Photoperiodism and Related Phenomena 

in Plants and Animals. (A.A.A.S., Washington, D. C.), pp. 601-624. 
PEPPER, H. J., 1937. Breaking dormancy in the sugar beet webworm, Loxostcge sticticalis L., 

by means of chemicals. /. Econ. Ent., 30: 380. 
SCHNEIDERMAN, H. A., AND C. M. WILLIAMS, 1953. The physiology of insect diapause. VII. 

The respiratory metabolism of the Cecropia silkworm during diapause and meta- 
morphosis. Biol. Bull, 105: 320-344. 
SIIAKHBAZOV, V. G., 1961. The reaction to the length of daylight and the light receptor of 

the pupa of the Chinese oak silkworm, Antheraca pern\i G. Doklady Akademii Nauk 

SSSR, (Biol.), 140: (AIBS translation) 249-252. 
STEEL, R. G. D., AND J. H. TORRIE, 1960. Principles and Procedures of Statistics. McGraw-Hill 

Book Co., Inc., New York. 
UMBREIT, W. M., R. H. BURRIS AND J. F. STAUFFER, 1957. Manometric Techniques and Tissue 

Metabolism. Burgess Publishing Co., Minneapolis, Minnesota. 
WIGGLESWORTH, V. B., 1934. The physiology of ecdysis in Rhodnius prolixus (Hemiptera). 

II Factors controlling moulting and 'metamorphosis.' Quart. J. Micr. Sci., 77: 191-223. 
DE WILDE, J., 1962. Photoperiodism in insects and mites. Ann. Rev. Entomol., 7: 1-26. 
DE WILDE, J., C. S. DUINTJER AND L. MOOK, 1959. Physiology of diapause in the adult Colorado 

beetle (Leptinotarsa decemlincata Say). I. The photoperiod as the controlling factor. 

/. Ins. Physiol., 3: 75-85. 
WILLIAMS, C. M., 1946. Physiology of insect diapause : The role of the brain in the production 

and termination of pupal dormancy in the giant silkworm, Plat \samia cecropia. Biol. 

Bull., 90: 234-243. 
WILLIAMS, C. M., 1947. Physiology of insect diapause. II. Interaction between pupal brain 

and prothoracic glands. Biol. Bull., 93: 89-98. 
WILLIAMS, C. M., 1948. Extrinsic control of morphogenesis as illustrated in the metamorphosis 

of insects. Groii'th Symposia, 12: 61-74. 



.. 

ON THE LUNAR ORIENTATION OF SANDHOPPERS 
(AMPHIPODA TALITRIDAE) 

F. PAPI AND L. PARDI 

Institute of General Biology, University of Pisa, Pisa, Italy, and 
Zoological Institute, University of Turin, Turin, Italy 

The sandhopper, Talitrus saltator, possesses a mechanism of astronomical ori- 
entation which enables it to return by the shortest route to its habitat (the moist 
area of the beach) when it has been taken away from it. During the day, the ani- 
mals use the position of the sun to orient (Pardi and Papi, 1952, 1953), while 
during the night they are able to orient by means of the moon (Papi and Pardi, 
1953, 1954, 1959; Papi, 1960). The oriented escape occurs when the animals are 
placed in unfavorable surrounding conditions. For example, a very low degree of 
relative humidity or an elevated temperature in the surroundings can act as re- 
leasing stimuli. The statistical study of oriented behavior during the return to the 
sea has been carried out by placing the animals in a concave glass observation 
chamber on whose dry walls the sandhoppers climb in the direction of the sea. 

As early as 1953 we were able to establish that the solar orientation is con- 
trolled by an endogenous rhythm and this was confirmed in further research (Papi, 
1955; Pardi and Grassi, 1955). More recent research on the lunar orientation 
(Papi and Pardi, 1959) indicated that a second rhythm, independent of the solar 
one, may control the angle of orientation with the moon. In fact, even animals 
kept in constant darkness from sunset onwards or from the new moon preceding 
the night of the experiment are, for the most part, capable of correct orientation. 

As soon as animals are introduced into the observation chamber they orient 
toward the moon. Only on particularly warm dry nights do the animals change 
spontaneously from this positive phototactic behavior to a correct orientation 
toward the sea. It is therefore necessary to heat the observation chamber or to 
dehydrate the air inside in order to bring about an oriented escape. In the experi- 
ments reported in 1959, as well as those in this report, we have always used the 
method of heating the base of the observation chamber. 

Enright has recently (1961) published a paper on the lunar orientation of 
Orchestoidea corniculata in which he arrived at conclusions which differed in part 
from our own. Enright experimented with animals held in three different sets of 
circumstances: (A) those kept in constant darkness for ten or more hours before 
the experiment ("constant darkness") ; (B) animals placed in constant darkness 
at the time of collection and re-exposed to natural light one hour prior to sunset 
and thus held until the moment of the experiment ("natural light") ; (C) animals 
treated like the former but redarkened two hours after moonrise ("redarkened"). 

While in the "natural light" and "redarkened" animals Enright found a correct 
orientation, in the "constant darkness" animals there seemed to be only the 
tendency to assume a constant angle with the moon, regardless of the lunar stage or 

97 



F. PAPI AND L. PARDI 

position. From this evidence the author drew the conclusion that the existence of 
a continuously operating endogenous lunar periodicity, which we postulated for 
Talitrus, is not admissible for Orchestoidca. Enright suggested that the correct 
orientation of the "natural light" and the "redarkened" animals could be explained 
(p. 155) as "a single-cycle night-time orientation rhythm re-initiated by the 
appropriate stimuli each night." 

Concerning the methods used by Enright, we may note that he neither heated the 
observation chamber nor dehydrated the air inside. In addition, he has observed 
that in his experimental conditions the repeated use of photo-bulb flashes at short 
intervals, for making photographic recordings, produced a considerable change in 
the angle of orientation and in the dispersion of the animals. In the first series of 
experiments, in each of which three photographs were made, the angle of orientation 
with the moon tended generally to increase, and there was a similar increase in the 
dispersion. In two successive experiments, in which ten photographs were made, 
the author observed instead that (p. 152) "the ultimate angle of orientation with 
the moon . . . was much smaller than the initial angle." 

Since the results of Enright's "constant darkness" animals contrast with those 
which we obtained from the sandhoppers maintained in darkness from sunset or 
from the new moon preceding the night of the experiment, and in view of his 
results on the effects of photo-bulb flashes, which throws doubt upon the validity 
of the method we employed, we wished to perform a new series of experiments on 
Talitrus saltator. 

MATERIALS AND METHODS 

We used Talitrus saltator Montagu from the beach at Castiglione della Pescaia 
(Grosseto) where the theoretical line of escape is 201. The apparatus was the 
same as that used for the experiments reported in 1959. The observation chamber 
was constantly heated, so that the air temperature of the interior was around 22 C. 

For the series of experiments, 1, 3, and 4, the animals had been collected the 
morning before the night of the experiment, between 1000 and 1115. Only for 
the series of experiments 2 was the collection made in the afternoon (18001845), 
about one hour before sunset. After collection the animals were kept in darkness, 
in jars containing moist sand, until the moment of their introduction into the ob- 
servation chamber. The introduction always took place in moonlight and without 
the aid of artificial light. After five minutes the first photograph was taken, and 
then eleven others were made at one-minute intervals, except in one case where 
only ten followed (experiment 4b). 

The experiments were performed with a perfectly clear sky on four nights in 
the summer of 1961. On each night two to four experiments were carried out with 
as many different groups of animals. For each experiment a variable number of 
animals (from 16 to over 50) was used, but the number of positions totally recorded 
(see Table I) is not always a multiple of the number of photographs taken, because 
of the fact that on some photographs some individuals were not visible (perhaps 
because the photographs were taken while the animals were jumping). Thus, we 
obtained single distributions from individual photographs and were able to cal- 
culate accumulated distribution. 

For each distribution (single or accumulated), we calculated the angular value 



LUNAR ORIENTATION OF SANDHOPPERS 



99 



of the average orientation direction (RD) and the length of the resultant vector, 
which measures the degree of scatter (OR) (see Pardi and Papi, 1953, p. 463, 
footnote 1). We have also calculated the average RD and OR for the single 
distributions, but they are so close to the RD and OR of the accumulated distribu- 
tion (a maximum difference of 1 and of 0.01, respectively), that we have not 
considered it necessary to report them. 

The expected, or theoretical, angle of orientation is the horizontal angle between 
the moon and the direction to the sea. The observed angle of orientation is the 
horizontal angle between the average direction of orientation (RD) and the moon. 
The position of the moon at the mid-point of each experiment was used. In de- 



TABLE I 

Orientation of Talitrus saltator with moon 













6. 






9. 


10. 






1. 






4. 




No. of 






Ob- 


Theo- 


11. 


12. 


ence 
num- 
ber 


2. 
Date 


3. 

Time 


Lunar 
stage : 
days 


Lunar 

azimuth 


re- 
corded 
posi- 


7. 
RD* 


RDi**- 
RD 


served 
angle 
with 


retical 
angle 
with 


Differ- 
ence. 
9-10 


Vector 
length 
(OR) 












tions 






moon 


moon 






la 


26 Au- 


21 h 53 m 30 3 


16 


128 


192 


177 


4- 2 


+49 


+ 73 


-24 


0.78 




gust 






















Ib 


27 Au- 


Ol h 37 m 30 s 


16 


189 


252 


178 


1 


-11 


+ 12 


-23 


0.79 




gust 






















Ic 


27 Au- 


04MO m 30 s 


16 


231 


228 


205 


- 5 


-26 


-30 


+ 4 


0.80 




gust 






















2a 


27 July 


22 h 06 m 30 3 


16 


144 


352 


165 


+ 2 


+21 


+57 


-36 


0.74 


2b 


28 July 


00 h 25 m 30 s 


16 


179 


341 


171 


- 3 


- 8 


4-22 


-30 


0.66 


3a 


28 July 


22 h 30 m 30 s 


17 


134 


240 


153 


-15 


+ 19 


4-67 


-48 


0.80 


3b 


29 July 


02 h 58 m 30 s 


17 


204 


191 


197 





- 7 


- 3 


- 4 


0.95 


4a 


3 July 


Ol h 15 m 30 s 


20 


126 


300 


157 


+ 8 


+31 


+ 75 


-44 


0.89 


4b 


3 July 


02 h 06 m s 


20 


137 


585 


150 


+ 6 


+ 13 


+64 


-51 


0.61 


4c 


3 July 


02 h 50 m 30 s 


20 


148 


192 


166 


+ 4 


+ 18 


+35 


-35 


0.85 


4d 


3 July 


03 h 23 m 30 s 


20 


157 


299 


168 


+ 12 


+ 11 


4-44 


7TO 

J\J 


0.68 



* RD : The resultant direction of orientation calculated by means of the complete distribution. 
** RDi: The resultant direction of orientation calculated from the registered distribution of 
the first photograph. 

termining these angles we have used for convenience a notation of positive when 
the moon was to the animal's left and negative when it was to their right. 

RESULTS 

7. Effects oj the photo-bulb flash 

In each experiment, the variations of the RD of the OR in the single distribu- 
tions, as compared with the RD and the OR of the accumulated distributions, are 
represented in Figures 1 and 2. The oscillations of RD are modest, with a differ- 
ence between maximum and minimum at the most less than 20. In one case (3a) 



100 



F. PAPI AND L. PARDI 



187 '/\ / x ^ a 188 

1 7 7 ? \ / ; n o 


/^ X *~* X ' 


167 


\/ -v 

\s*'* 168- 


..._._.^ 



215 
205 


.1C 175 J /x 23 

A/ V ' 4 / V \ 

- 1 G "% v * ^ 






^-' x 


*\_/^ - / 


\ 


1 9 5j * 1 5 5 - 


^ 


,,,J 


2b 163 o| 


3a 




^*v 


"*""*""-. .-- 


171 


. . / 1-1 






^ \./ ^V- X " J 




w ~ 

161J 143- 





V 


207- 


3b 167 , 


>x \ 



107.1 


^-. ./* N - 
x -=*.^ * 1 ; 7 


^*"* x ^. 



187 



160 



I.-. 



i 

o 

150 


\ 


\s- 


o 




140- 





176 



166 



156 



178? 



168 



158' 



\ 



\ 



FIGURE 1. The resultant directions (RD) in each experiment. The horizontal line in 
each graph indicates the value of the resultant direction calculated by means of the accumulated 
distribution ; circles indicate the RD of the single distributions. 



LUNAR ORIENTATION OF SANDHOPPERS 



101 



0,88 


0.89 

^~*v *~~m- 
r \ / ft 7 Q 


V\ 

\ 


,78 - 
0.68- 


\ / \ / "' J 

A/ 

0,69 




0,90 


^ ... ^ C 0,8^ 


A?A. 


n on . 


/ *~* \ Q7L 


^/ 


U,oU 

j 


\ ' ] 




0.70 | 


"^^ ' 


\ / 


0,76- 

Dec . 


2 b 090 . 


3a 


0,56- 


.^- xl 

a7o 


,., 


095 J 


3b 099 

i -* _ ^ _A^"^** n p Q 


^ Aa 


U,33 

0,90 


V 

0,79 


* V 


0,71; 

061 


4b a95 . 

/^ .>.--n* Q85 


1 4c 

i ... 


0,51 


^.99 ,*** 

^ 

0,75 




0.78 ' 

Oc o . 


\ /\ /A 




>DO 


V *^v" * 




0,58 







FIGURE 2. The degree of scatter (OR) in each experiment. The horizontal line in each 
graph indicates the OR calculated by means of the accumulated distribution ; circles indicate the 
OR of the single distributions. 



102 



F. PAPI AND L. PARD1 



there is an oscillation of 29, but it seems clear that the animals, during the first 
three photographs, still showed the positive phototactic behavior which precedes 
the correct orientation (the lunar azimuth was 134). It may be noted that the 
variations of the RD do not seem to follow a fixed pattern. The differences be- 
tween the RD of the first single distribution and the RD of the accumulated dis- 
tribution are also shown in Table I. The algebraic average of these differences is 

less than 1 (X = +054' + 7 12') and consequently we retained, as already 
stated, the use of the RD of the accumulated distribution as a valid index of the 
animals' orientation. 

The degree of the scatter (Fig. 2) does not vary considerably, nor in any regu- 
lar pattern. It should also be noted that the OR is always greater than the value 
of 0.50 which Enright considered the minimal value having significance when 
calculated for groups of 20 or 30 animals. 

2. The variations of the angle of orientation 

In general the animals seemed to be well oriented, since only in two experiments 
out of eleven did the RD of the accumulated distribution differ more than 45 
from the theoretically expected direction. Moreover, the degree of scatter, as 
we have seen, was always so small that the distributions must be considered 



+ 80 








4a 


U 




+ 60 






03a 

U 4b| 










f 



D2a 
Ac 






4-40 
+20 
0- 






4d 










2b 












1b 























-20 




3b 























-40 


1C 












-20 +20 +40 +60 +8 



FIGURE 3. Correlation between the theoretical angle of orientation (ordinate) 
and the observed angle of orientation (abscissa). 



LUNAR ORIENTATION OF SANDHOPPERS 



103 



significant. In comparison with animals of this same population tested immediately 
after capture (Papi and Pardi, 1959, p. 587), the results of the current experiments 
show a stronger tendency to deviate towards south and southeast, a tendency which 
is generally more marked the nearer the moon is to the east. 

A fact of critical importance is whether the angle of orientation with respect to 
the moon varies in relation to the moon's position or tends instead to oscillate 
randomly around some other value. Figure 3 represents the correlation between 
the observed angles and the theoretical angles. The graph indicates a good corre- 
lation by inspection. The coefficient of correlation is r = 0.83 (t r = 4,462; 
P < 0.01). The observed angles of orientation are positive in the first part of 
the night, tend to become smaller, and then negative, as the moon approaches the 
west (Table I, col. 9 and Fig. 4). Thus, there is a regular variation in the 
angles, except for the one night of 3 July. We have not noticed any difference in 



26 August 
21 h 53 m 30 S 





27 August 
04 h 10 m 30 S 



128 



205 

FIGURE 4. An example of the variation of the angle of orientation with moon shown by 
two lots of animals in the course of the same night. Each number inside the circle shows the 
number of positions recorded in that sector. The open circles outside represent the moon ; 
the arrows indicate the resultant directions. 

the orientation or in the dispersion between animals collected in the morning ( series 
of experiments 1, 3, and 4) and in the afternoon (series of experiments 2). 

DISCUSSION 

From the results of the above mentioned experiments, we may observe that: 
(' 1 ) the repeated use of the photo-bulb did not produce any noticeable disturbances ; 
(2) the animals continued regularly to vary the angle of orientation, even when 
they were not exposed to natural light variations resulting from the sunset and 
from the moonrise. 

The stronger tendency to deviate towards the south and southeast, in com- 
parison to our former experiments in which animals were tested immediately after 
capture, may be attributed to the interruption in the natural light cycle. This 
could have induced a small shift in the endogenous mechanism of lunar orientation 
without, however, having arrested its functioning. 



104 F. PAPI AND L. PARDI 

We are able, therefore, to confirm the validity of our previous results on 
Talitrus saltator (1953, 1959) and do not find, for this animal, support for Enright's 
hypothesis that the sunset and/or the moonrise can start (p. 155) "a single cycle of 
appropriately time-compensated lunar orientation." The hypothesis of an en- 
dogenous lunar periodicity, which operates continuously, still seems the most 
plausible. 

The differences between our results and those of Enright could be due to 
different mechanisms of orientation in the two species. We think it more likely, 
however, that they are due to the fact that since Enright did not heat the observa- 
tion chamber nor dehydrate the air inside, the releasing stimuli on certain nights 
did not attain the necessary threshold. It should be noted that all the experiments 
with animals "not kept in constant darkness" (Enright's Table II) were made in 
a single night (6-7 August) and that on the same night even animals "kept in 
constant darkness" (Table I, last five experiments) oriented with nearly the same 
degree of precision. It is therefore probable that new experiments on Orchestoidea 
could explain the difference between Enright's results and ours. 

The research has been supported by grants from the Rockefeller Foundation. 
The authors wish to express sincere thanks to Dr. H. E. Savely for careful reading 
of the manuscript, and for his assistance in the translation. 

SUMMARY 

1. Under the action of appropriate releasing stimuli the amphipod, Talitrus 
saltator, is capable of orienting by the position of the moon in a relatively constant 
azimuth. This ability still functions after 10 or more hours of captivity in constant 
darkness. 

2. The direction of escape and the degree of scatter are not noticeably influenced 
by the repeated photo-bulb flashes. 

3. Additional evidence supports the hypothesis that the lunar orientation of 
Talitrus is due to a continuously-operating lunar physiological rhythm. The hy- 
pothesis of a single-cycle night-time orientation rhythm, put forward for Or- 
chestoidea corniculata, does not seem applicable to Talitrus saltator. 

LITERATURE CITED 

ENRIGHT, J. T., 1961. Lunar orientation of Orchestoidea corniculata Stout (Amphipoda). 

Biol. Bull, 120: 148-156. 
PAPI, F., 1955. Experiments on the sense of time in Talitrus saltator (Montagu) (Crustacea- 

Amphipoda). Expcrientia, 11: 201. 
PAPI, F., 1960. Orientation by night: the moon. Cold Spring Harbor Syinp. Quant. Biol., 25: 

475-480. 
PAPI, F., AND L. PARDI, 1953. Ricerche suH'orientamento di Talitrus saltator (Montagu) 

(Crustacea- Amphipoda). II. Sui fattori che regolano la variazione dell'angolo di 

orientamento nel corso del giorno. L'orientamento di notte. L'orientamento diurno di 

altre popolazioni. Zeitschr. vergl. Physiol., 35: 490-518. 
PAPI, F., AND L. PARDI, 1954. La luna come fattore di orientamento degli animali. Boll. 1st. 

Mus. Zool. Univ. Torino, 4: 1-4. 



LUNAR ORIENTATION OF SANDHOPPERS 105 

PAPI, F., AND L. PARDI, 1959. Nuovi reperti sull'orientamento lunare di Talitrus saltator 

Montagu ( Crustacea- Amphipoda). Zeitschr. vergl. Physiol., 41 : 583-596. 
PARDI, L., AND M. GRASSI, 1955. Experimental modification of Direction-finding in Talitrus 

saltator (Montagu) and Talorchestia deshayesci (Aud.) (Crustacea- Amphipoda). Ex- 

perientia, 11 : 202-210. 
PARDI, L., AND F. PAPI, 1952. Die Sonne als Kompass bei Talitrus saltator (Montagu) (Am- 

phipoda-Talitridae). Naturzviss., 39: 262-263. 
PARDI, L., AND F. PAPI, 1953. Ricerche sull'orientamento di Talitrus saltator (Montagu) (Crus- 

tacea-Amphipoda). I. L'orientamento durante il giorno in una popolazione del litorale 

tirrenico. Zeitschr. vergl. Physiol., 35: 459-489. 



Vol. 124, No. 2 April, 1963 

THE 

BIOLOGICAL BULLETIN 

PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY 



THE EFFECT OF P-CHLOROMERCURIBENZOATE ON AMOEBOID 
MOVEMENT, FLAGELLAR MOVEMENT AND GLIDING MOVEMENT 

SHIGEMI ABE 
Department of Biology, Faculty of Science, Osaka University, Osaka, Japan 

It has been shown (Abe, 1959, 1963) that the rotational protoplasmic streaming 
observed in plant cells is dependent upon sulfhydryl (-SH) groups in the proto- 
plasm. In addition, De Robertis and Peluffo (1951) reported an intimate relation 
between the motility of a flagellated bacterium, Proteus vulgaris, and -SH groups. 
Therefore, it seemed worthwhile to investigate the relationship between -SH groups 
and other types of protoplasmic motility. Amoeboid movement, ciliary or flagellar 
movement and gliding movement have been dealt with in this report. 

MATERIALS AND METHODS 

Amoeboid movement: A strain of Amoeba proteus was chosen as an example 
of an organism exhibiting amoeboid movement. This strain was found to be 
abundant in cultures of Elodca, and was cultivated by the addition of some uncooked 
rice grains. The specimens were prepared in the following manner. 

The amoebae were sucked up, together with the original culture medium, in a 
pipette and dropped onto the center of a glass slide. In order to avoid mechanical 
pressure from the coverslip, a pair of thin glass rods about 10 mm. in length and 
0.5 mm. in diameter were laid on the slide. Then, the coverslip was put on the 
drop carefully. Excessive water was removed with a piece of filter paper. 

Flagellar movement: Flagellated cells of coenobia of Pandorina morum served 
as a material exhibiting flagellar movement. The material was collected from a rice 
paddy in a suburb of Osaka. 

As these organisms move freely and actively through an aqueous medium, the 
preparation used for the amoebae is impracticable for their observation. In this 
case, only one glass rod was used as a support. A coverslip was brought in position 
after a drop of suspension of coenobia was carefully placed at the center of the 
slide. A piece of filter paper was next brought in contact with the edge of the 
coverslip resting on the glass slide, so that only the suspension medium would be 
sucked off. In this procedure, coenobia in the suspension were carried along with 
the stream of water and were caught and held in the wedge-shaped space between 

107 
Copyright 1963, by the Marine Biological Laboratory 



108 SHIGEMI ABE 

the coverslip and slide. By this means, we obtained a preparation with a slightly 
slanted coverslip, under which locomotion was prevented, but with which active 
movement of the flagella could be observed. Whenever exchange of the medium 
was required, a solution was made to flow in the same direction in order not to 
refloat the coverslip and not to release the trapped organisms. 

Gliding movement: A species of Oscillatoria with a comparatively large diameter 
(about 16 /A) and a species of pinnated marine diatom. Nitzschia longissima, were 
used as materials. The cells of Oscillatoria used in the present study are blue-green 
and disk-shaped, and the ratio of the diameter to the length is about 7. Each fila- 
ment tip of this species is straight. 

The filaments of Oscillatoria are well known to show a bending oscillatory motion, 
but when a short filament (300 p.- 400 /* in length) is isolated free from others, this 
motion is transformed into one of gliding along the axis of the filament, so that the 
speed of motion becomes measurable. The speed and direction of the movement 
of the filament, however, undergo changes even under constant environmental 
conditions, and temporary cessation often takes place during the movement. There- 
fore, in order to compare the behavior of the filament under a particular experi- 
mental condition with the control, continuous measurements were made for as long 
as 2-4 hours. 

As the motion of Nitsschia is slow and takes a fairly straight path, it is not 
difficult to determine the changes in velocity of locomotion in this organism with 
an ocular micrometer and stopwatch. 

Reagents used throughout this investigation were commercial p-chloromercuri- 
benzoate (PCMB) and L-cysteine supplied from Wako Pure Chemical Industries, 
Ltd. Since PCMB is only slightly soluble in acid or neutral solutions, it was at 
first dissolved in solutions of sodium hydroxide and subsequently neutralized with 
hydrochloric acid. In the present experiments, the sodium concentration was kept 
at W~' 2 M, 10 3 M and 10' 4 M according to the concentrations of PCMB. Conse- 
quently, for control solutions, 10~ 2 M, 10~ 3 M and 10~ 4 M NaCl solutions were 
employed. The sodium content of cysteine solutions was made equal to that of the 
corresponding control solutions. For the marine diatoms, Van't Hoff artificial sea 
water was used for a basal solution, and solutions of the above reagents were made 
to contain a concentration of each ion equivalent to that of the artificial sea water. 

All of these experiments were performed at room temperature. 

EXPERIMENTAL RESULTS 
Part I 

Experiments were first carried out with solutions of PCMB at different con- 
centrations. Before each experiment, the medium surrounding the organism was 
removed with a piece of filter paper, and either saline (in cases of Amoeba, Pan- 
dorina and Oscillatoria) or artificial sea water (in the case of Nitsschia) was 
introduced. After having observed the behavior of the organisms under this condi- 
tion, the solutions were replaced with PCMB solutions containing a certain amount 
of NaCl or equivalent amount of salts to the Yan't Hoff artificial sea water. Con- 
centrations of the PCMB solutions used were lO' 3 M, 10" 4 M and lO' 5 M. The 
results obtained are summarized in Table I. Detailed data are given in the following 
text. 



EFFECT OF PCMB ON CELL MOTION 

TABLE I 

The effects of PCMB on amoeboid movement, flagellar movement 
and gliding movement at different concentrations 



Material 


Concentrations of PCMB 


10-3 m 


10-J M 


10--' \t 


Amoeba 
Pandorina 
Oscillatoria 

Nitzschia 


cytolyzed after 10 sec. 
stopped within several sec. 
stopped within ,? niin. 


cytolyzed after 8 niin. 
stopped after 2 niin. 
stopped after 10 niin. 
stopped after 1 niin. 


temporary cessation 
no effect 
no effect 
stopped in 7 niin. 



Amoeba 

(10~ 3 M) : Immediately after the application of a control solution 10~ 2 M 
NaCl -movement of Amoeba became sluggish and the pseudopodia were drawn 
in. But in a short time, the pseudopodia were re-formed and within 2-3 minutes 
the organism regained its normal state completely. 

After complete recovery, NaCl solution was replaced with 10~ 3 M PCMB con- 
taining 10~- il/ sodium. Immediately thereafter the boundary between the cell and 
the external medium became obscure, and the contents of the organism were 
extruded in about 10 seconds. 

(10"* M) : When the culture medium was exchanged with 10~ 3 M NaCl, 
movement became at first a little sluggish, but it soon recovered. About 30 seconds 
after admitting 10~ 4 M PCMB, the cell gradually became inactive and all pseudo- 
podia were drawn in. Later the organisms became spherical and after 8 minutes 
they cytolyzed. 

(10 ' M) : With 10' 5 M PCMB the effect was much less pronounced. As 
a control, a solution 10~ 4 M NaCl was first applied in this case, but there was 
no observable effect. The saline was then replaced with 10~ 5 M PCMB containing 
10~ 4 M sodium. Within 40 seconds all pseudopodia were retracted and locomotion 
ceased, but in a short time tiny pseudopodia reappeared. In three minutes the 
motion recovered completely and the organism survived in this medium without 
any sign of pathological changes. 

Pandorina 

("10~ 3 M) : With 10"- M NaCl, the flagellar beating of the coenobium remained 
normal. When the saline was exchanged with 10~ 3 M PCMB containing 10~ 2 M 
sodium, flagellar beating came to a standstill in several seconds. Sometimes flagella, 
which had been beating, were detached from the cell bodies at their basal ends after 
several seconds. 

( 10~ 4 M) : The effect of PCMB on flagellar beating was still to be seen even 
at a concentration of 10~ 4 A/; the beating ceased within two minutes. 

(10- 5 M ') : PCMB solution at a concentration of 10 -"' M had no visible effect 
on flagellar beating. 



110 



SHIGEMI ABE 



u. 
o 

fc 








-2 

TIME IN MINUTES 

FIGURE 1. The effect of PCMB on the gliding movement of Oscillatoria at different con- 
centrations. From to 40 minutes, materials were in NaCl solutions at 10~ 2 M (O O), 10~ 3 M 
( ) and 10~ 4 M (O O) ; and at 40 minutes, NaCl solutions were replaced with 10~ 3 M 
PCMB + lO- 2 M sodium (O O), 1Q- 4 M PCMB + 1Q- 3 M sodium ( ) and with 1Q- 5 M 
PCMB + 10- 4 M sodium (O O). 

Oscillatoria 

(10~ 3 M) : As done above, Oscillatoria were first placed in NaCl solutions 
which served as the control. The motion of the filaments, however, was not in- 
fluenced in NaCl at the concentration of 10' 2 M or less. When 10~ 2 M NaCl 
was replaced with 10~ 3 M PCMB containing 10~ 2 M sodium, the motion of the 
filament slowed down and stopped within three minutes and no resumption of 
motion occurred in the same medium within 20 hours. 

(10' 4 M') : With 10" 4 M PCMB, the motion decreased, gradually resulting 
in a complete standstill after 10 minutes. There was no visible change in cell 
morphology over the observation period of 18 hours. 

(10" 5 M) : With 10^ 5 M PCMB, the gliding motion of the filament was no 
longer affected at all. These results are illustrated in Figure 1. 

Nitsschia 

(10 4 M) : The artificial sea water medium was replaced with the artificial sea 
water containing PCMB. In the plain medium, no change was observable. About 




30 



40 



-10 - 



TIME IN MINUTES 



FIGURE 2. The effect of PCMB on the gliding movement of Nitzschia at different concen- 
trations. From to 20 minutes, cells were placed in Van't Hoff artificial sea water, and at 20 
minutes, medium was replaced with the artificial sea water containing 10~ 4 M (O O) and 
10-5 M ( ) PCMB. 



EFFECT OF PCMB ON CELL MOTION 



111 



one minute after the replacement with 10 4 M PCMB the gliding motion of the 
organism fell into complete cessation with no spontaneous recovery (Fig. 2). 

(1O 5 M) : Even with 10~ r ' M PCMB, the motion became sluggish and stopped 
completely in seven minutes (Fig. 2). 

Part II 

Once the inhibiting concentrations were determined, experiments were carried 
out to verify the hypothesis that the effect of PCMB was a specific one on sulfhydryl 
groups. The plan of the experiments with three materials Amoeba, Pandorina 
and Oscillatoria is illustrated in Table II. 

Details of the results of these experiments are given below for each material, 
the paragraph numbers corresponding to the numbers heading the columns in 
Table II. 

TABLE II 



A general outline of the procedure followed and the results obtained 
in the present experiments. See the text for details 



o 



<LI 



H 
Z 

w 

H 
U 

H 



(1) 
SALINE 



PCMB 

inhibition 



CYSTEINE 

recovery 



(2) 
SALINE 



PCMB 

inhibition 



SALINE 
no recovery 



(3) 
SALINE 



PCMB 

+ CYSTEINE 
no inhibition 



(4) 
SALINE 



CYSTEINE 

no visible effect 



Amoeba 

(1) About four minutes after treatment with 10~* M PCMB, this solution was 
replaced with 10~ 2 M cysteine containing 10~ 3 M sodium. Within 60 minutes, no 
visible change was observed and the cells remained spherical. In such cells, Brown- 
ian motion of granules was observed. Then about 70 minutes after replacement, 
a short pseudopodium was extended, although no conspicuous flow of cytoplasm 
was observed. After 100 minutes, a pseudopodium with pronounced endoplasmic 
streaming was formed. By about 160 minutes, complete recovery of amoeboid 
movement had occurred. 

(2) As the control of the former experiment in item (1), PCMB solution was 
replaced with plain saline instead of a cysteine solution. When 10~* M PCMB 
solution was replaced, about four minutes after its application with 10~ 3 M NaCl, 
2-3 huge vacuoles appeared in the cells. Thirty minutes later, the cells became 
spherical, and disintegrated in about 40 minutes. 

(3) PCMB and cysteine solutions were mixed before they were applied to the 
organisms. In this mixed solution, when the amount of the latter exceeds that of 
the former, the SH combining power of PCMB was expected to vanish. Control 
solutions 10~ 3 M NaCl were replaced with a mixed solution containing 5 X 



112 



SHIGEMI ABE 



10-' M PCMB, 5 >: 10-' 1 M cysteine and 10~ 3 M sodium. Under these conditions, 
the organisms behaved quite normally, and there was no observable effect. In 
support of this statement it is to be added that the amoebae often phagocystosed 
ciliated cells in this same medium. 

For the sake of contrast, a solution containing 5 X 10~ 5 M PCMB and 10 3 M 
sodium was used. About two minutes after application, movement of the organisms 
almost stopped except for the Brownian motion of the granules inside. Three 
minutes later, Brownian motion stopped ; six minutes later, the organism became 
spherical and giant vacuoles appeared; 25 minutes later, the organism began to 
cytolyze. 

Pandorina 

(1) Immediately after the complete cessation of the flagellar beating with 10^ 3 M 
PCMB, the inhibitor solution was replaced with 10~ 2 M cysteine containing 10~ 2 M 
sodium. Then, recovery of the flagellar beating was observed within several min- 
utes. No abnormal behavior was to be seen later in the organisms which had 
recovered. Even in the coenobium from which flagella had been partially lost by 
PCMB, flagella left attached recovered their normal beating. 

(2) No recovery of flagellar beating took place when 10~ 3 M PCMB was re- 
placed with 10 - M NaCl solution. 

(3) Materials were treated with a mixed solution of 5 X 10' 4 M PCMB, 
5 X 10~ 3 M cysteine and 10~ 2 M sodium. With this solution, there was no visible 
effect on flagellar beating. 

On the other hand, when treated with a solution containing 5 X 10~ 4 M PCMB 
and 10 ~- M sodium, beating stopped within a period of 20^1-0 seconds and, some- 
times, the flagella separated from cell bodies. 

(4) Simple application of 10~ 2 M cysteine solution with 10" 2 M sodium normal 
coenobia exerted no effect. 




TIME IN MINUTES 



FIGURE 3. Inhibition of the gliding movement of Oscillatoria with PCMB and restoration 
with cysteine. From to 12 minutes, materials were immersed in 10~ 3 M NaCl and from 12 
to 25 minutes in 10-* M PCMB containing 10~ 3 M sodium, at 25 minutes, PCMB solution was 
replaced with 10~ 2 M cysteine containing 10~ 3 M sodium (O O) or 10^ 3 M NaCl ( ). 



EFFECT OF PCMB ON CELL MOTION 113 

Oscillatoria 

(1) A1)out 10 minutes after the complete cessation of gliding motion with 
10~ 4 M PCMB, the inhibitor solution was replaced with 10~ 2 M cysteine containing 
10~ 3 M sodium. Although in this case a complete recovery was not attained, a 
partial recovery was observed (Fig. 3). 

(2) No sign of recovery was seen in the organism when 10~ 3 M NaCl was 
substituted for 10~ 4 M PCMB (Fig. 3). 

(3) A mixed solution of 5 X 10- M PCMB, 5 X KH M cysteine and 10~ 3 M 
sodium produced no effect, either on motion or on any other aspect of behavior of 
the organisms. 

In contrast to the above, a solution containing 5 X 10~ 5 M PCMB and 10~ 3 M 
sodium and no cysteine stopped the gliding motion of the filament within fifteen 
minutes. 

(4) In solutions containing 10~ 2 M cysteine and 10~ 3 M sodium, there was no 
visible effect on normal organisms. 

DISCUSSION 

On the basis of the foregoing experiments, it may be concluded that the effect 
of PCMB on the protoplasmic motility of different types is more or less identical : 
First, motion becomes sluggish ; this is then followed by complete cessation at some 
particular concentration. Second, motion is restored completely or partially through 
subsequent application of cysteine in concentrations higher than those of PCMB 
applied. Third, when the reactivity of PCMB upon SH groups has been elimi- 
nated in advance, through addition of cysteine, PCMB exerts no effect on the 
cellular motion. 

It is well known that PCMB reacts with SH groups to yield mercaptides. 
According to Olcott and Fraenkel-Conrat (1947) and Barron (1951), PCMB is 
the most advantageous of all the known SH reagents in the following respects: 
PCMB reacts with -SH groups with high specificity and reacts with no other 
protein groups; its combination with SH groups can be thoroughly dissociated 
with the addition of the reagents having -SH groups in their molecules; these 
reactions are carried out under physiological conditions. Therefore, it is reason- 
able to conclude that the cessation and recovery of the movements dealt with in 
the present paper are closely related to the blocking and liberation of SH groups 
in the protoplasm. 

In the author's previous work, it was amply verified that the rotational proto- 
plasmic streaming in plant cells has close relation to SH groups in the protoplasm. 
In addition, DeRobertis and Peluffo (1951) applied sulfhydryl reagents to the 
cells of a flagellated bacterium, Proteus vulgaris, and reported that the movement 
of the cells became sluggish and stopped completely under the influence of these- 
reagents, especially PCMB. Recovery of the movement was brought about with' 
subsquent application of cysteine or glutathione, but simple washing with saline or 
buffer solution did not remove the action of the inhibitors. These results were' 
interpreted by the authors as evidence of the presence of essential -SH groups 
in the contractile protein of bacterial flagella, or in enzymes especially involved in 
the mechanism of flagellar motion. 



114 SHIGEMI ABE 

Singer and Barren (1944) showed that PCMB completely inactivated the 
ATP-ase function of myosin, and subsequent treatment with glutathione restored 
full activity. Bailey and Perry (1947) have found in addition that the SH groups 
of myosin are essential for actomyosin formation. These results point out the 
significance of SH groups in muscular contraction. Whether or not the inhibi- 
tion by PCMB of cellular motions is also due to suppression of ATP-ase activity 
of a contractile protein is still unknown. But the importance of the SH groups 
in the basic mechanism of a variety of protoplasmic motions seems now well 
established. 



The author wishes to express his most cordial thanks to Professor N. Kamiya 
of Osaka University for his helpful advice and directions throughout this investi- 
gation, and to Professor R. D. Allen (Princeton University) for the kind help 
in preparation of the manuscript. 

SUMMARY 

1. The effect of p-chloromercuribenzoate (PCMB), a specific sulfhydryl re- 
agent, upon several types of cellular motion amoeboid movement, flagellar move- 
ment and gliding movement was studied. 

2. The effect of PCMB on the motions of different types is more or less iden- 
tical : First, motion becomes sluggish ; this is then followed by complete cessation 
at some particular concentration. Second, motion is restored completely or partially 
through subsequent application of cysteine in concentrations higher than those of 
the PCMB previously applied. Third, when the reactivity of PCMB upon SH 
groups has been eliminated with cysteine in advance, PCMB exerts no effect on 
the cellular motion. 

3. From the fact that PCMB is bound to SH groups with high specificity, 
it is reasonable to conclude that the cessation and recovery of the cellular motion 
are closely related to the blocking and liberation of -SH groups. 

LITERATURE CITED 

ABE, S., 1959. Rotational protoplasmic streaming and SH. Kagaku (Science), 29: 361-362 

(in Japanese). 
ABE, S., 1963. The effect of p-chloromercuribenzoate on rotational protoplasmic streaming 

in plant cells. In press. 
BAILEY, K., AND S. V. PERRY, 1947. Role of sulfhydryl groups in the interaction of myosin 

and actin. Biochcm. Biophys. Acta, 1 : 506-516. 

BARRON, E. S. G., 1951. Thiol groups of biological importance. Adv. Enzyiuol., 11 : 201-266. 
DE ROBERTIS, E., AND C. A. PELUFFO, 1951. Chemical stimulation and inhibition of bacterial 

motility studied with a new method. Proc. Soc. E.rp. Biol. Med., 78: 584-589. 
QLCUTT, H. S. AND H. FRAENKEL-CONRAT, 1947. Specific group reagents for proteins. Chan. 

Rev., 41 : 151-197. 
SINGER, T. P., AND E. S. G. BARRON, 1944. Effect of sulfhydryl reagents on adenosinetri- 

phosphatase activity of myosin. Proc. Soc. E.rp. Biol. Med., 56 : 120-124. 



FERTILIZATION IN PECTINARIA (=CISTENIDES) GOULDII 

C. R. AUSTIN i 
Marine Biological Lahumfory, Woods Hole, Muss. 

Pectinaria gouldii was selected for study because the animal was readily avail- 
able and its egg is more suitable for phase-contrast microscopy that those of many 
other marine invertebrates, owing to the lack of interfering cytoplasmic structures 
such as refractile droplets or coarse granulations. The observations are published 
because, with the exception of Tweedell's (1962) report on cytoplasmic inclusions, 
only brief notes on Pectinaria eggs have appeared in the literature (Wilson. 1936; 
Costello et al., 1957; Tweedell, 1959, 1960, 1961), and because fertilization in these 
eggs has several features of interest. 

METHODS 

The animals were collected in the Woods Hole area during June and July, 
1961, and maintained in the laboratory in fingerbowls containing sand and sup- 
plied with running sea water. Gametes were obtained by removing an individual 
to a separate dish and carefully breaking down the conical tube in which the animal 
lives, starting from the thin end. As a rule, shedding occurred during this process 
and it was mostly these gametes that were used in the studies to be described. 
More rigorous methods were often needed to provoke shedding, such as prodding 
or pinching the animal with forceps, but gametes thus obtained were in the main 
found to be unsatisfactory, most of the eggs being incompletely grown and the 
spermatozoa unlikely to leave their packets (Fig. 1) and develop free motility. 

Eggs were examined by phase-contrast microscopy, both in the fresh state and 
after fixation and staining as whole-mounts. A solution of 10% glacial acetic 
acid in absolute alcohol was used for fixation ; staining was effected with aceto- 
carmine (Schneider's). Some eggs in the fresh state were examined by fluo- 
rescence microscopy; they were treated with acridine orange solution ( 0.0005 % 
in sea water) and ultraviolet radiation. Eggs after semination were also fixed in 
10% formalin solution, embedded in paraffin, sectioned at 6 ^ and stained with 
haematoxylin and eosin. 

OBSERVATIONS 
Primary oocyte 

The Pectinaria egg took the form of a concavo-convex disc 50-60 ^ in diameter 
and 25-30 /A in thickness. It was bounded by a closely applied transparent vitel- 
line membrane about 2 ^ thick. The cytoplasm contained numerous finely granular 
elements and was devoid of large lipoidal droplets. 

1 Present address : Physiological Laboratory, Downing Street, Cambridge, England. 

115 



116 



C. R. AUSTIN 




All photographs were taken by phase-contrast microscopy and are reproduced at X 1300. 
FIGURE 1. Sperm packets. (Fresh material.) 

FIGURE 2. Primary oocyte with intact germinal vesicle. A spermatozoon had recently 
entered the egg (at top). (Fresh material.) 



FERTILIZATION IN PECTINARIA 117 

The germinal vesicle (Fig. 2) was large relative to the size of the egg, its 
diameter being about half that of the vitellus. , A single nucleolus was usually pres- 
ent, in addition to a number of scattered chromosomes. In the living egg, the 
chromosomes were only vaguely distinguished but they became more evident when 
they condensed just before breakdown of the germinal vesicle. In the condensed 
phase of the fixed egg, they could clearly be seen to have the forms characteristic 
of bivalents. The attempt was made to count the chromosomes in 26 such eggs ; 
estimates ranged from 17 to 21, with a mean of 19.2, uncertainty being due par- 
ticularly to the presence of several very small chromosomes. 

Primary oocytes treated with acridine orange solution and ultraviolet radiation 
exhibited bright green fluorescence in a zone around the nucleolus and in numerous 
cytoplasmic granules, and bright red fluorescence in about an equal number of cyto- 
plasmic granules. The nucleolus itself, the nucleoplasm and the hyaline cytoplasm 
did not have noticeable fluorescence. 

Oocytes examined histologically displayed a moderate cytoplasmic basophilia, 
especially in the granular elements, and a strong basophilia in chromosomes and 
in the band of material surrounding the nucleolus. 

Maturation 

. The egg was always shed as a primary oocyte ; maturation changes began almost 
immediately, the proportion of eggs showing the changes varying between 10% 
and 90% in different individuals. The chromosomes condensed, though remaining 
scattered, the nuclear membrane became irregular in outline and disappeared, and 
the chromosomes aggregated in the center of the egg as a loose group which became 
the metaphase plate of the first polar spindle (Fig. 4). Further progress was 
seen only after sperm penetration ; when this occurred, meiosis was resumed and 
the first and second polar bodies were successively separated. The polar bodies 
were always extruded into the middle of the concave side of the egg (Fig. 6). 
The first polar body had a tendency to divide partially or completely, and thus 
three polar bodies were occasionally seen. Eggs could be left in sea \vater for at 
least three hours after shedding without losing their capacity to proceed, on semina- 
tion, with apparently normal fertilization. 

FIGURE 3. A similar egg to that in Figure 2, showing germinal vesicle (with two or 
three bivalents in approximate focus ) , a sperm head in the cytoplasm and a residual fertiliza- 
tion cone. (Acetic alcohol and acetocarmine.) 

FIGURE 4. Primary oocyte showing polar view of the first meiotic metaphase plate. 
(Acetic alcohol and acetocarmine.) 

FIGURE 5. A similar egg to that in Figure 4, showing a recently penetrated sperm head 
which had lost its refractility and become stained. (Acetic alcohol and acetocarmiiu-. ) 

FIGURE 6. Secondary oocyte; surface view showing polar bodies. (Fresh material.) 

FIGURE 7. Ootid with male and female pronuclei approaching full development. (Fresh 
material. ) 

FIGURE 8. Ootid with two male pronuclei which appear fused, but were in fact sepa- 
rated by their nuclear membranes and smaller female pronucleus. (Fresh material.) 

FIGURES 9-12. Successive stages of pronucleus development in an egg in which polar body 
emission failed and two female pronuclei were formed. Seemingly intact nuclear membranes 
still separated the pronuclei in Figure 12 which give the appearance of having fused. (Fresh 
material.) 



C. R. AUSTIN 



Sperm penetration 



The frequency with which sperm packets hroke up to liherate free and actively 
motile spermatozoa differed much between different animals. With good material, 
complete break-up of the packets occurred within 5 or 10 minutes. 

Spermatozoa were occasionally seen to become attached to the vitelline mem- 
brane and then pass gradually through this structure ; actual entry into the vitellus 
was not witnessed, the spermatozoa under observation all being stopped apparently 
at its surface. Many instances were encountered in which a spermatozoon had 
just entered the egg cytoplasm, and generally the sperm head lay within a small 
fertilization cone. There was no evidence of elevation of a fertilization membrane. 
Penetration could apparently occur at any point on the egg surface except in the 
concave region. A little further progress could be watched until the sperm head 
became indistinguishable among the cytoplasmic granules. Subsequent stages were 
studied in fixed and stained eggs, and it was evident that the sperm head enlarged 
and lost its retractility as it advanced into the egg (Fig. 5). The course of events 
of entry into eggs before the breakdown of the germinal vesicle was broadly the 
same as that into eggs of later stages, except that in the former case rather larger 
fertilization cones were often formed and the sperm head remained refractile and 
did not enlarge (Figs. 2 and 3). Fertilization proceeded only in eggs penetrated 
after the germinal vesicle had started to break down. Allowing eggs to stand in 
sea water for three hours after shedding and before semination did not affect the 
frequency of sperm penetration. 

Pronucleus development 

During its early stages of development, the female pronucleus could clearly be 
seen to be made up of separate portions, karyomeres (Fig. 10), which later came 
together and fused to form a single nucleus. Evidence of septa could also be seen 
in the early male pronucleus (Fig. 9). Residual walls or strands tended to persist 
in both pronuclei at later stages, and seemed to be responsible for irregularities 
in the outlines of the pronuclei which often looked roughly polygonal (Figs. 7, 8 
and 12). At full development, male and female pronuclei came into close apposi- 
tion and sometimes gave the impression of having undergone fusion (Fig. 12) ; 
in no instance, however, could it be said with certainty that actual fusion had 
occurred, for there always seemed to be a thin membranous structure separating 
the two nuclei. The chromosome groups deriving from the male and female pro- 
nuclei in the prophase of the first cleavage division were regularly found to be 
quite separate. 

Observations indicated that the male pronucleus was formed almost always 
before the female ; it was possible for successful fertilization to be initiated by 
sperm penetration occurring as soon as the germinal vesicle had broken down, 
and thus the male pronucleus was formed during the early maturation changes in 
many eggs. Male pronucleus development then seemed to be suspended until 
female pronucleus development was under way, for intermediate and later stages 
of growth were regularly found to be synchronous in the two pronuclei. 

In several trinuclear eggs that had two polar bodies (presumptive polyspermic 
eggs), two pronuclei were seen to be equal in size and both were larger than the 



FERTILIZATION IN PECTINARIA 119 

third (Fig. 8) ; in a trinuclear egg that had no polar hody (in which it was sur- 
mised that two female pronuclei had developed from the egg chromosomes after 
anaphase separation), two pronuclei were smaller than the third (Figs. 9-12). 
The inference is that in this species the male pronucleus is larger than the female. 
The pronuclei contained several small nucleoli (Fig. 7). 

Anomalies 

The chief anomalies noted were: (1) polyspermy, (2) refertilization, (3) pos- 
sible early gynogenesis, and (4) failure of polar-body formation. There were no 
instances of androgenesis or of spontaneous development beyond the first polar 
metaphase (rudimentary parthenogenesis). 

Polyspermy. The occurrence of polyspermy was inferred, in early fertilization, 
from the presence of two or more sperm heads in the vitellus, or, at later stages, 
from the presence of three or more pronuclei in eggs having two (or three) polar 
bodies (Fig. 8). A total of 163 polyspermic eggs was recorded; one was tetra- 
spermic, fourteen trispermic and the rest dispermic. The tetraspermic egg and 
all but one of the trispermic eggs had intact germinal vesicles. Some eggs ob- 
served in stages of the first cleavage division had multipolar spindles ; in six, the 
spindle \vas tripolar, and in two tetrapolar. These eggs all had two polar bodies 
and were also assumed to have undergone polyspermic fertilization. 

Refertilisation. The term is used to denote the entry of a second (or third) 
spermatozoon into the vitellus much later than the first (or first two), so that the 
male elements are clearly in different stages of development. The distinction is 
drawn with polyspermic fertilization in which the male elements are in the same 
stage of development and apparently advance synchronously. Altogether, ten 
examples of refertilization were seen, in each of which there was a sperm head 
that had not undergone noticeable change. Four of these eggs had in addition 
late-stage male and female pronuclei, two showed an early male pronucleus with 
the egg chromosomes in the second polar prophase, two had early male pronuclei 
and egg chromosomes in second polar metaphase, one egg was in the metaphase 
of the first cleavage division and one egg was two-cell. 

Gynogenesis. Possible examples of early gynogenesis were presented by three 
eggs ; each had a well formed female pronucleus and two polar bodies, but the 
penetrating sperm heads had not appreciably changed. 

Failure of polar-body formation. Two eggs were seen in the course of fertiliza- 
tion, both of which lacked polar bodies ; they were kept under continuous obser- 
vation while the pronuclei formed. In one, there were two presumptive female 
pronuclei and a larger male pronucleus (Figs. 9-12). In the other the sperm head 
changed into an early male pronucleus, and the egg chromosomes, having failed to 
undergo anaphase separation, became incorporated into a single female pronucleus, 
which was about the same size as the male pronucleus. 

Time relations and incidence of features of fertilization 

The time relations of fertilization were studied in four experiments, the results 
of which were reasonably consistent ; details of the most elaborate experiment are 



120 C. R. AUSTIN 

set out in Table I. From these sets of data, the times for fertilization stages were 
estimated to he as follows : Breakdown of the germinal vesicle became evident a 
few minutes after shedding and most eggs showed these changes by 10 minutes. 
The first polar division had advanced to metaphase in a few eggs by 20 minutes 
and most eggs had reached ( or, if penetrated, had passed ) this stage within the 
first hour. Sperm penetration began immediately after semination and virtually 
all the eggs that were destined to be penetrated contained spermatozoa by 10 min- 
utes after semination. Most of the eggs that were penetrated and lacked the 
germinal vesicle exhibited first polar bodies and early male pronuclei between 20 
and 25 minutes after semination. Second polar bodies and female pronuclei made 
their appearance within the next 10 minutes. The whole pronuclear phase of 
fertilization lasted about half an hour, condensation of chromosomes in the pro- 
phase of the first cleavage division being evident in most eggs between 50 and 60 
minutes after semination. Mitosis seemed to pause in metaphase, and cytoplasmic 
cleavage was completed in most eggs 20 to 30 minutes later, namely about 80 
minutes after semination. 

The following inferences are drawn from the observations summarized in Table 
I. Some eggs fail to undergo any maturation changes, even after standing in sea 
water for an hour or more : the data show that the proportion of oocytes with 
intact germinal vesicles fell to 51% (23% + 20% + 8%) at 45 minutes after 
shedding and suffered only a small reduction thereafter. At each interval from 
45 minutes onwards, it was found that roughly half the unchanged eggs had been 
penetrated by spermatozoa ; evidently, sperm penetration neither provoked nor 
inhibited maturation. The frequency of sperm penetration into primary oocytes, 
whether before or during maturation, did not change significantly from that ob- 
served at 10 minutes after semination this was despite the presence, for well over 
an hour, of many active spermatozoa in the medium. The proportion of eggs 
penetrated was somewhat higher in those with maturation changes (67% ) than 
in those without ( 54%. ) . On the other hand, the incidence of polyspermy was 
much higher in eggs with intact germinal vesicle (24% of penetrated eggs) than 
in eggs at later stages (4% of penetrated eggs). 

DISCUSSION 

The limits of the breeding season of Pcctinoria at Woods Hole have not yet 
been determined, but Costello ct al. (1957) noted that ripe animals could be secured 
at least during August. The present investigation was carried out on animals 
collected in June and July, which may have been before the peak of the season. 
This would explain why many animals could not be induced to shed, why sperma- 
tozoa obtained from others often failed to become free-swimming, and why occa- 
sionally as few as 10% of oocytes underwent maturation. The relatively low rate 
of sperm penetration that occurred (around 60%} despite the presence of many 
surplus free-swimming spermatozoa may have been owing also to unripeness of 
animals providing the gametes. 

According to Costello et al. ( 1957 ) , polar bodies are separated about 29 min- 
utes after semination, and the first cleavage occurs at about 54 minutes. The 
corresponding times recorded in the present series were 30-35 minutes and about 



FERTILIZATION IN PECTINARIA 



121 



=0 

. 

^e, 





6 

s 



-S 

t>3 

X. 
*J 



t^ 
8 

S 

S 

Ci] R 

-j 
=: ^ 

.< ^3 
j 



8 

' <?* 

s 







e 

s 



o 
3 





S 

^ 





o 






















| 


* 






^~, 














1 


*^~ 

01 






:t 














o 


CN 






o 














a, 


00 






"" 














o 


* 






* 












en 




-f 


















ho 


i_ 








vC 










0) 


W 


V 
C 


<x 






J^ 










00 

C- 




ai 


NO 






T 














fl. 








NO 














o 


i 






O 








01 








A 






A 
















f ^ 


^ 




















^ ^ 


^ 


~_^ 




-s5 


o 






87 


o 


1/3 


CN * i rf 


o 


T 1 1 1 O 




01 d 




O) 


00 


LO 


i>o 


* 


^ ^ x ^-*' 


^^^^ 


S ^ ^^ ^ ^ *" 




^~*^ ^~** 


^ *~ 


^"^ 


01 


^ ' 


" 


^ H 


^ ro rsj 





O ^ ~- 


C 


OO " 


ON 


f^. 


^ 


-t 






oo o ~- 




LO 






ON. 






ON 






o' ^^ 


_^^ 


O 


LO 


r^^ 


NO" 






01 


o 


^ 


* T i r- 


^ o 


0) O O 


c 


~ d 


01 




>~e 




N 


00 


"* ' s ' ^~^ 


^ ' - ^ 


* ^ ^^-^ - 


^ ^ 


- - - - 


" ' 




00 


' 






^ t 1 ^- LO 


r^ H ^^ 


OO OJ ^H 


oi c 


OO 01 


^H 




re 


ON 






^O ^O f^>l 




r- 




-i- 


O 






re 
















^^ 






01 


* 




OO S - -. 
~ 01 LO 


o -^ d 


So'*? 

c 


S 2- 


^ O 








^? 




E 


ON LO oo 


-^ LO i 


O\ i-H ^^' 


Tt" Ol 


^ ^ 






re 


01 






LO NO ^^ 




LO 




1 






















1 








01 






^^ -H rr; 


LO LO 

o, ^ 


00 S"^ 


^e d 










i 


vC 


fM 


^^ ^ ^-^ 


^~^ 












* ' 


' ^ 






re -t 1 ** 


O) LO O 


ON f- " 


r e ^~~* 








r o 


LO 






LO -f -H 




LO re 


























*" ' 










04 






^ S"^ 


oT^ 


x s ON -~o 












^ 






01 01 OO 


1 f T- 1 


ON ^ 










J^ 


LO 


^' 


O 


- ' ^^ 














(T3 


^ 






-f r^ O 


ON 1^ 01 


Ol rf Ol 










CN 


01 






LO -^ O) 


01 -H 


01 Tf 


















00 


g 




















NO 


re 


01 










.^ 









^ ' 


^ ' 


"" 










o 








o 


01 


LO 










01 








* 


NO 




















00^ 


2 












01 




10 


















o 








LO 


t^ 












01 








(-H 


01 




















"c5 S 


<u P P 
^ C C 


1/3 C 


E E 


S E 










d 


C 

1 


u- i 

QJ O 

E c. a. 
zj <f t & 


g oj o 

"n & a 
u< in to 


S.S.SL 

rt t/3 U5 


S.S, 

tn in 


ui cn 










'g 




U ~ -H ^-1 



















M 

C 


C 

o 


j= 9 A 


1-. 

o A 


E A 


A 


A 








OX) 
bfl 


3 

T) 


rt 

C 

E 


1 '' + + 


J+ + 


^+ + 


+ + 


+ + 








QJ 


en 


05 


















rt 


u 


^ 


















u 


13 


4-t 

"rt 


> & 

u. >, 






^2 *-* 








So 
bfl 




1 


E 


E 


2 ^-' 






o' 


"C 


t _ t 


, 




_- 


H 


H 


E ^ 








o o 


M 

c 


"aJ 
u 


"aJ 







4J 

^c 













c^ 


O 


01 


i 


r- 







<u 

c 
- 



u 



in 

:/. 
;/. 

O 



K) 

OJ 

U 



OJ 
bC 



0) 
U 



tn 

(U 



~ 



- 

5 B | 
1- 



<U 



, * 



C. R. AUSTIN 

80 minutes. The authors just cited asserted also that prormclear fusion took place 

at about 40 minutes ; in the present series, the early prophase changes of the first 

cleavage division were observed between 50 and 60 minutes after semination. 

pThe implication is that fertilization was slower in the present series, probably 

lowing to differences in water temperature (24 C. in the study of Costello et al. r 

\18 C. in the present series). 

Two features having temporal association with the breakdown of the germinal 
vesicle are worthy of comment, and they may be attributable to the same under- 
lying cause. These features are the large differences in the incidence of polyspermy 
before (24%} and after (4%) germinal vesicle disappearance, and the difference 
in the fate of penetrating sperm heads before and after the event. The difference 
in the incidence of polyspermy was the more striking since eggs with intact germinal 
vesicle showed a lower frequency of sperm penetration (54%) than did those 
undergoing maturation (67%}. The deficiency in the former group presumably 
lay in the block to polyspermy, and the inference is that, when maturation begins, 
the nature of the reaction shown by the egg plasma membrane to sperm contact 
undergoes an important change. Normally, sperm penetration involves early 
fusion between egg and sperm plasma membranes (Colwin and Colwin, 1960; 
Szollosi and Ris, 1960; Friedmann, 1962), and it may well be that the block to 
polyspermy is provoked by such fusion. However, in oocytes before maturation, 
sperm entry may possibly occur in the way it was classically supposed to happen, 
namely, by a process resembling phagocytosis. With a phagocytic form of sperm 
absorption, the spermatozoon would be engulfed with its membrane intact and 
surrounded at a short distance by an envelope of egg plasma membrane. Early 
stages of what could be phagocytic engulf ment of a sperm by a Lytechinus primary 
oocyte have been described by Franklin and Metz (1962). Under these circum- 
stances, the sperm head would be shielded from cytoplasmic agents normally 
responsible for bringing about its metamorphosis into a male pronucleus. Thus, 
the high incidence of polyspermy and the persistence of unchanged sperm heads 
in the oocyte before breakdown of the germinal vesicle can both be explained by 
the assertion that the reactivity of the egg plasma membrane to sperm contact is 
not yet of the kind needed for normal fertilization. 

The fluorescent colors displayed by oocytes treated with acridine orange and 
ultraviolet radiation conform in general to the descriptions given by Tweedell 
(1959, 1960, 1962). They are difficult to interpret. In fixed mammalian tissue 
subjected to the same treatment, yellow and red fluorescence was shown to be 
associated with the presence of DNA and RNA, respectively (Armstrong, 1956). 
Living mammalian eggs were found to have strongly green and red fluorescent 
structures ; here the green color was regularly evident in DNA-containing elements 
but the distribution of the red color was consistent with an association, not with 
RNA, but with mononucleotides (Austin and Bishop, 1959). In Pectinaria, the 
.red fluorescence could reasonably be ascribed to the presence of mononucleotides, 
but the cytoplasmic green fluorescence seems most unlikely to be indicative of DNA. 
The cytoplasmic DNA of Paracentrotits eggs is not demonstrable by histochemical 
methods, owing apparently to its low order of polymerization ( Hoff- Jjzfrgensen, 
1954), and the acridine-orange-induced green fluorescence in the cytoplasm of 
Arbacia oocytes was not removed by DNAase (H. Esper, 1962, personal com- 



FERTILIZATION IN PECTINARIA 123 

munication ) . Possibly, green fluorescence in these marine eggs can be ascribed 
to component proteins having an appropriate degree of polymerization. 



The work described in this report was done when the author, a member of the 
External Scientific Staff, Medical Research Council, London, was F. R. Lillie Me- 
morial Fellow for 1961 at the Marine Biological Laboratory, Woods Hole, Massa- 
chusetts. The observations on induced fluorescence were made in collaboration with 
Dr. C. B. Metz who also provided the fluorescence microscope and other facilities 
required for this procedure. 

SUMMARY 

1. The eggs shed by Pectinaria gouldii were in the stage of primary oocytes 
with intact germinal vesicles. Maturation began promptly, but proceeded only to 
the first metaphase, which some eggs reached in 20 minutes, and further progress 
depended upon sperm penetration. 

2. Spermatozoa entered eggs rather more readily after than before breakdown 
of the germinal vesicle, and only those that entered after the event developed into 
pronuclei. It is suggested that spermatozoa entering eggs before maturation may 
be engulfed in a manner resembling phagocytosis, as distinct from membrane fusion. 
Polar bodies were extruded about 20-25 minutes and 30-35 minutes after semi- 
nation at 18 C. Male pronuclei were evidently larger than the female. Early 
prophase of the first cleavage mitosis was seen between 50 and 60 minutes after 
semination, and cell division took place 20 to 30 minutes later at 18 C. 

3. The incidence of polyspermy observed was 24% before and 4% after break- 
down of the germinal vesicle. Eggs undergoing refertilization, possible early 
gynogenesis. and development after failure of polar-body formation were also seen. 

4. Oocytes treated with acridine orange displayed green and red cytoplasmic 
granules, as well as a green zone around the nucleolus. The red fluorescence may 
have denoted the presence of mononucleotides ; the green fluorescence was not 
considered specific to DNA. 

LITERATURE CITED 

ARMSTRONG, J. A., 1956. Histochemical differentiation of nucleic acids by means of induced 

fluorescence. Ex p. Cell Research, 11: 640-643. 
AUSTIN, C. R., AND M. W. H. BISHOP, 1959. Differential fluorescence in living rat eggs 

treated with acridine orange. E.rp. Cell Research, 17: 35-43. 
COLWIN, A. L., AND L. H. COLWIN, 1960. Changes in the spermatozoon during fertilization in 

Hydroides hc.vagomts (Annelida). II. Incorporation with the egg. /. Biophvs. 

Biochem. Cytol, 10: 255-274. 
COSTELLO, D. P., M. E. DAVIDSON, A. EGGERS, M. H. Fox AND C. HENLEY, 1957. Methods for 

Obtaining and Handling Marine Eggs and Embryos. Marine Biological Laboratory,. 

Woods Hole, Massachusetts. 
FRANKLIN, L. E., AND C. B. METZ, 1962. Electron microscope study of sperm entry into sea 

urchin oocytes. Biol. Bull., 123 : 473. 
FRIEDMANN, L, 1962. Cell membrane fusion and the fertilization mechanism in plants and 

animals. Science, 136: 711-712. 



124 C. R. AUSTIN 

HOFF-J0RGENSEN, E., 1954. Deoxynucleic acid in some gametes and embryos. In: Recent 
Developments in Cell Physiology, p. 79. Ed. by J. A. Kitching. Butterworths Scien- 
tific Publications, London. 

SZOLLOSI, D. G., AND H. Ris, 1960. Observations on sperm penetration in the rat. /. Bioplivs. 
Biochem. Cytol., 10: 275-283. 

TWEEDELL, K. S., 1959. Induced fluorescence in marine eggs. Biol. Bull.. 117: 429-430. 

TWEEDELL, K. S., 1960. Differential staining of cytoplasmic inclusions in eggs of Pectinaria. 
Biol. Bull., 119: 345. 

TWEEDELL, K. S., 1961. Factors affecting germinal vesicle breakdown in Pcctinurici (Cis- 
t en ides) go u Id ii. Biol. Bull.. 121 : 412. 

TWEEDELL, K. S., 1962. Cytological studies during germinal vesicle breakdown of Pectinaria 
goitldii with vital dyes, centrifugation and fluorescence microscopy. Rial. Bull.. 123 : 
424-449. 

WILSON, D. P., 1936. Notes on the early stages of two polychaetes, Xcplttliys hoinbergi 
Lamarck and Pectinaria koreni Malmgren. /. Mar. Biol. Assoc., 21 : 305-310. 




THE RELATION BETWEEN INTENSITY OF INDUCTOR 

AND TYPE OF CELLULAR DIFFERENTIATION OF 

RANA PIPIENS PRESUMPTIVE EPIDERMIS 1 - 

LESTER G. EARTH AND LUCENA J. EARTH 

Marine Biological Laboratory, Woods Hole. Mass, and the Departments of Zooloi/y 
of Columbia University and Barnard Collci/e. New } ark 27, N. ) . 

In a recent paper (Earth and Earth. 1962), we suggested that lithium chloride 
applied in various concentrations might he expected to induce the presumptive 
epidermis to differentiate into various cell types. Relatively low concentrations 
of lithium chloride regularly induced nerve, higher concentrations, pigment cells 
and. occasionally, very high concentrations for long periods induced muscle cells. 

The present investigations are concerned with a more detailed study of the 
effects of lithium chloride upon the presumptive epidermis of Rana pipiens gastrula. 

EXPERIMENTAL PROCEDURE 

An important departure from the methods used in previous experiments is 
the omission of agar as a substrate during the period of reaggregation of the 
partially dissociated cells of the presumptive epidermis. The aggregates are now 
prepared on the glass surface of a medium-sized stender dish and transferred 
at desired intervals to the small-sized stender dishes used as culture dishes. The 
use of a glass surface rather than agar became desirable when it was found that 
agar has a neuralizing effect on presumptive epidermis. This effect was first 
found in experiments which called for a short period during which the aggregates 
remained on agar before being transferred to the glass-bottom culture dishes. 
If the contact between the aggregate and agar was less than 30 minutes, the 
aggregates differentiated into a sheet of epithelium with ciliated patches instead 
of radial nerve with oriented cells. Aggregates exposed to agar from 45 minutes 
to 21 hours regularly differentiated with some type of nerve cells present. 

The evidence for the neuralizing effect of agar is presented in Tables I and II. 
The first experiment compares the cellular differentiation obtained after the 
aggregates have been exposed to glass, to agar, and to glass surrounded by 
agar for 5 hours. The "glass surrounded by agar" was achieved by covering 
the bottom of the dish with agar in the usual manner and then removing strips 
of agar and placing the aggregates on the glass surface so exposed. 

When no agar is present in the dish the aggregates upon transfer to culture 
dishes spread out into a thin sheet of epithelium with ciliated patches. Only one 
aggregate formed a few nerve fibers. In contrast the aggregates which had 
been exposed to agar. with or without direct contact, differentiated into spreading 

1 Aided by a grant from the National Institutes of Health ( RG-3322 ) . 

2 The authors acknowledge with gratitude the valuable assistance of Iris Nelson. 

125 



126 



LESTER G. EARTH AND LUCENA J. EARTH 



TABLE I 

The neuralizing effect of agar upon presumptive epidermis 

Stage 11 gastrulae. Operating and culture medium: solution with 1 mg./ml. globulin (Bios). 
Agar: washed repeatedly in distilled water over prolonged period at 3-5 C. Dissolved in stand- 
ard solution. Conditions: operation, dissociation in Versene and preparation of aggregates 
carried out on agar or glass as indicated and aggregates left for period indicated before transfer 
to small culture dishes without agar. 



Exp. No. 


Conditions 


Time, hrs. 


No. of 
aggregates 


Cellular differentiation 


1 


Glass surface 


5 


30 


Epithelium 




Agar surface 


5 


30 


Spreading nerve 




Glass surface, agar sides 


5 


45 


Spreading nerve 


2 


Agar film 


7 


40 


Epithelium 






23 


35 


Epithelium 


3 


Agar, pH 6.9 


6-8 


75 


Radial nerve 




Agar, pH 8.3 


6-8 


75 


Radial nerve 




Agar, pH 7.0 


0.5 


35 


Epithelium 




Agar, pH 7.0 


2.5 


40 


Radial nerve 




Agar, pH 7.0 


6.0 


35 


Radial nerve 



nerve which is characterized by a network of nerve fibers and very few epithelial 
cells. 

The neuralizing effect of agar is thus not the result of the type of surface 
provided by agar but is probably brought about by a diffusible substance present 
in the agar. Repeated washing with cold distilled water over prolonged periods 

TABLE II 

Effect of nitrogen and oxygen on the neuralizing action of agar 
Time: time in hours of exposure to agar and gas 



Time, hrs. 


N 2 


2 


Air 


No. 


Diff. 


No. 


Diff. 


No. 


Diff. 


0.1 


40 


Epithelium 


30 


Ciliated masses 


40 


Epithelium 


0.5 


85 


Epithelium 


30 


Epithelium 


40 


Epithelium 


0.75 


35 


Epithelium 


50 


Short nerve 


40 


Short nerve 


1.0 


50 


Epithelium 


40 


Short nerve 


35 


Short nerve 


1.5 


115 


Epithelium 


75 


Radial nerve 


35 


Radial nerve 


1.75 










40 


Radial nerve 


2.25 










35 


Short nerve 


3.0 


45 


Epithelium 






40 


Short nerve 






Radial nerve 










3.3 


35 


Radial nerve 










4.0 






35 


Radial nerve 






5.0 






15 


Radial nerve 


40 


Spreading nerve 


6.0 










35 


Radial nerve 


7.0 










40 


Spreading nerve 


7.75 


75 


Spreading nerve 






40 


Spreading nerve 






Radial nerve 










8.0 


25 


Spreading nerve 


35 


Spreading nerve 






21.0 










35 


Radial nerve 



INDUCTOR INTENSITY 127 

did not remove this substance. Since streptomycin sulfatc antagonizes the neural 
differentiation produced by agar, we tried washing the agar with this compound. 
Such treated agar still induced nerve, however. 

The second experiment recorded in Table I deals with a special situation 
in which an extremely thin film of agar was obtained by pouring hot agar into 
a hot dish and immediately pouring out the excess. A continuous film of agar 
resulted upon cooling, and the aggregates exposed to such a film for 7 to 23 hours 
differentiated into epithelial sheets with no nerve present. We conclude that 
the agar surface is not the neuralizing agent. 

The third experiment was prompted by the possibility that, in the preparation 
of the agar by boiling in our salt solution, the phosphates precipitated and the 
pH became elevated. Other experiments show that sodium bicarbonate added to 
the solution to bring the pH to 8.8 induces nerve. The data show that agar at 
pH 6.9, 8.3 and 7.0 induces nerve if the exposure is more than 30 minutes. Since 
our salt solution for culture of the cells is at pH 8.0-8.2 and does not induce 
nerve at these pH's, the neuralization by agar dissolved in the salt solution is not 
brought about by a change in pH. 

Table II gives additional evidence for the neuralizing effect of agar. Aggre- 
gates are permitted to form on agar and are left there for varying lengths of time. 
In air and in oxygen. 45 minutes' exposure is sufficient for neuralization. Further 
exposure results in more and more neural differentiation with fewer and fewer 
epithelial cells present. No significant differences in differentiation occurred in 
oxygen as compared with air. In nitrogen, however, epithelial sheets with little 
or no nerve were obtained for as long as 90 minutes. After 3.3 hours' exposure 
to agar and nitrogen, differentiation was predominantly nerve in character. 

As a result of the experiments recorded in Tables I and II, we decided to 
carry out all the operations on glass for the experiments presented in this paper. 
Otherwise the method is the same as outlined in Earth and Earth (1962). 

CLASSIFICATION OF CELLULAR DIFFERENTIATION 

We will use the same designation of the types of differentiation obtained by 
treatment with various compounds as was used in Earth and Earth (1962). 
A brief description follows : 

1 . Epithelium : The cells spread out in a sheet of epithelial cells, patches of 
which become ciliated. A variant results when the aggregate fails to attach 
and free-swimming ciliated masses of cells result. 

2. Radial nerve : Some of the cells spread out as an epithelial sheet while the 
cells in a centrally located mass send out nerve fibers over the surface of the 
epithelial sheet. Later the cells in the epithelial sheet become elongated and 
oriented in parallel fashion about the central mass of nerve cells. The epithelial 
cells do not form cilia. 

3. Spreading nerve : Few epithelial cells are present and these few are 
scattered about the periphery of a diffuse nerve network. The latter is character- 
ized by an extensive migration of neuroblasts to form several loci for formation 
of nerve fibers. 



128 



LESTER G. BARTH AND LUCENA J. EARTH 







2 
V 




FiGi'RK 1. An "astrocyle" from a culture prepared by treating presumptive epidermis 
cells of stage 11 gastrula with a solution of lithium chloride (0.74 mg./ml. of standard solution) 
for 47 hours at 23 C. On return to standard solution the cells differentiated as large pigment 
ring cells and transformed into "astrocytes." Fixed and stained after 37 days of culture at 
23-24 C. 



INDUCTOR INTENSITY 129 

4. Short nerve : No epithelial cells are present. The neuroblasts remain in a 
dense mass and send out numerous nerve fibers which turn and run parallel to 
the mass of neuroblasts. 

5. Pigment cells : The original aggregate spreads out and the cells appear 
to repel each other so that single cells, approximately equidistant from each other, 
result. Most of these cells develop a ring of slate-gray pigment granules in 
contact with the nuclear membrane. These cells later form melanophores. Some 
mesenchyme cells also are present. 

In addition three new types of cellular differentiation have been obtained. 

6. "Astrocytes" : Large star-shaped cells with a large pigment ring and 
filaments containing granules. These cells persist for three to four weeks and 
resemble astrocytes. Figure 2 shows such a cell with its filaments attached to 
adjacent cells. Figure 1 shows a typical "astrocyte." 

7. "Neuroglia" : Very large cells with long thick filaments filled with granules. 
They first arise as large single cells and as the filaments develop they connect 
with the filaments of adjacent cells. The result is a network or sometimes a 
lattice work of cell filaments. The cells resemble neuroglia cells. Figures 3 
and 4 show these cells at 4 days while Figures 5 and 6 illustrate the connections 
between cells on the seventh dav. 

J 

8. Calcium-induced nerve : This type of cellular differentiation is not described 
by either radial nerve or spreading nerve or short nerve. Although there are 
some epithelial cells and these become oriented as in radial nerve, there are also 
many nerve fibers growing out from the central mass into a periphery relatively 
free of epithelial cells. These fibers resemble short nerve. Then, too, although 
there is some migration of neuroblasts as in spreading nerve, most of the neuro- 
blasts remain in a dense central mass. For purposes of identification we will 
use the term calcium-induced nerve to describe the above cultures (Figs. 7 and 8). 

Although any one of these types of cellular differentiation may be induced 
with high frequency, in any one experiment, there is almost always some other 
type present. For example after certain treatments with lithium chloride most 
of the aggregates will form pigment ring cells but a few aggregates will also form 
nerve cells in addition to pigment ring cells. In the tallies such a result would 
be classified as "pigment cells." 

Similarly, our controls are now epithelial sheets with ciliated patches but 
a few aggregates in some experiments have formed some nerve. These are 
still classified as "epithelium." 

In addition there are some inconsistencies in the tables. These are probably 
the result of one or all of three variables which are not accurately controlled. 
First, there is the stage of the donor gastrula. While we try to use stage 11 
consistently, stage 1 1 itself is a variable and of course the age of the presumptive 
epidermis is important in relation to its competence to be induced. Second, we 

FIGURE 2. Same as Figure 1 and illustrates tendency of "astrocytes" to make contact 
with adjacent cells. 

FIGURE 3. A 4-day culture of "neuroglia" cells, produced by culturing the presumptive 
epidermis of stage 11 gastrula in a solution containing 0.47 mg. of lithium chloride per ml. 
of standard solution. These cells contain no pigment rings. 

FIGURE 4. Same as Figure 3 but higher magnification, showing absence of pigment rings 
and presence of yolk. 



130 



LESTER G. EARTH AND LUCENA J. EARTH 




FIGURE 5. A 7-day culture of "neuroglia" cells prepared by culturing the presumptive 
epidermis of stage 11 gastrula in a solution containing 0.56 mg. of lithium chloride per ml. 
of standard solution. Most of the cells are connected with each other by protoplasmic filaments. 

FIGURE 6. A detail of Figure 5, showing the granular nature of the protoplasmic filaments. 



INDUCTOR INTENSITY 131 

have not accurately controlled the temperature during the treatment of the 
aggregates. We use room temperature which is controlled by air conditioners. 
In early experiments we found the vibration of constant temperature incubators 
to have a deleterious effect on the attachment of the aggregates. 

Finally the size of the aggregate we use varies and size may be important 
in determining the concentration of the test solution at the cell surfaces. For 
example, lithium chloride in low concentrations induces nerve cells while in higher 
concentrations pigment cells form. In a relatively large aggregate in a high 
concentration of lithium we find that the peripheral cells form pigment cells while 
the centrally located cells become nerve cells. Is this a result of the relatively 
large mass having been subjected to a gradient of concentrations of lithium, high 
at the periphery and low in the center? 

RESULTS 

We will first report the results of experiments designed to test whether induc- 
tors such as sodium bicarbonate, magnesium sulfate, ammonium sulfate and 
calcium chloride induce the presumptive epidermis to form various types of nerve 
when the neutralizing effect of agar is eliminated. Second, the results of an 
extensive series of experiments in which lithium chloride in various concentrations 
is used as an inductor for varying lengths of time will be presented. 

Sodium bicarbonate. The addition of 1 mg./ml. of NaHCCX, induces the 
presumptive epidermis to form nerve cells. Table III, Exp. 1, records the 
results of exposing the aggregates for periods varying from 0.1 hour to con- 
tinuous exposure. Actually some of the aggregates have been exposed longer 
than the recorded time since it takes about 20 minutes to prepare them. This 
probably accounts for the little nerve present with a short exposure of 0.1 hour. 
Typically, however, a short exposure results in extensive sheets of epithelial cells, 
some of which are ciliated. Nine hours' treatment induces extensive radial nerve 
together with elongate cells oriented in parallel fashion. Some of the aggregates 
form short nerve with no epithelial cells present. With longer exposure or con- 
tinuous exposure the aggregates exhibit no migration of epithelial cells but remain 
as a dense mass of cells. Nerve fibers develop slowly and turn to run parallel 
to the circumference of the mass of neuroblasts. Many fibers form a dense 
intricate pattern of fibers classified as short nerve. 

Exp. 2 confirms the first experiment and was designed to test the period of 
competence of the presumptive epidermis for neuralization by sodium bicarbonate. 
Again a short exposure results in a little nerve while the control shows no nerve. 
The competence after 0.5 hour remains unaltered and short nerve develops in 
sodium bicarbonate. After 7 hours' treatment with sodium bicarbonate the 
aggregates form short nerve in our standard salt solution. If, however, the 
aggregates are prepared in a standard salt solution and remain in it for 7 hours 
and then are transferred to sodium bicarbonate, little or no nerve develops and 
the cultures resemble controls very closely. Both sets of cultures contain pre- 

FIGURE 7. Calcium-induced nerve produced by culturing the presumptive epidermis of 
stage 11 gastrula in a solution containing 0.54 mg. CaCl 2 -2H 2 O per ml. of standard solution. 
Fixed at 7 days. 

FIGURE 8. A detail of Figure 7, showing high frequency of nerve fibers. 



132 



LESTER G. EARTH AND LUCENA J. BARTH 



dominantly epithelium. Thus, stage 11 presumptive epidermis loses its compe- 
tence to react to added sodium hicarhonate in 7 hours at 24 C. By this time the 
whole gastrula controls are in stage 12. 

Exp. 3 records the effects of a lower concentration of sodium bicarbonate. 
A sojourn of 0.5 hour in a concentration of 0.95 mg./ml. has no effect and the 
presumptive epidermis does not form nerve. If the aggregates remain in standard 
salt solution for 0.5 hour and then are cultured in 0.95 mg./ml., nerve is induced. 
Competence decreases after 7 hours at 19 C. and very little nerve is induced 
by added sodium bicarbonate. 

TABLE III 

Effects of sodium bicarbonate on the differentiation of presumptive epidermal cells, stage 11 
The normal concentration of NaHCOs in our standard solution is 0.2 mg./ml. In this and 
the succeeding tables the table headings are to be interpreted as follows. Stage: Shumway 
(1940) ; Treatment, cone. : the final concentration of the test substance in milligrams per milliliter 
of standard solution; Time: the length of time in hours during which the aggregates are exposed 
to the test solution; Culture cone.: the concentration of the test substance in milligrams per 
milliliter of the standard solution to which the aggregates are transferred; No.: the number of 
aggregates treated; Types of cellular differentiation: as defined in text. 





Treatment 


Culture 


Results 


No. 


Cone., 
mg./ml. 


pH 


Time, 
hrs. 


Temp., 
C. 


Cone., 
mg./ml. 


pH 


No. 
aggregates 


Types of cellular differentiation 




1.2 


8.9 


0.1 


21 


0.2 


8.0 


30 


Epithelium, little nerve 




1.2 


8.9 


9.0 


21 


0.2 


8.0 


35 


Radial nerve, short nerve 


1 


1.2 


8.9 


22.0 


21-25 


0.2 


8.0 


40 


Short nerve 




1.2 


8.9 


0.5 


24 


1.2 


8.9 


25 


Short nerve 




1.2 


8.9 


0.5 


24 


0.2 


8.2 


25 


Epithelium, little nerve 




0.2 


8.2 


0.5 


24 


0.2 


8.2 


25 


Epithelium, no nerve 




0.2 


8.2 


0.5 


24 


1.2 


8.9 


25 


Short nerve 


2 


1.2 


8.9 


7.0 


24 


0.2 


8.2 


25 


Radial nerve 




0.2 


8.2 . 


7.0 


24 


0.2 


8.2 


25 


Epithelium 




0.2 


8.2 


7.0 


24 


1.2 


8.9 


25 


Epithelium, little nerve 




0.95 


8.8 


0.5 


19 


0.2 


8.1 


35 


Epithelium, ciliated masses 




0.2 


8.1 


0.5 


19 


0.95 


8.8 


40 


Short nerve 


3 


0.95 


8.8 


7.0 


19 


0.2 


8.1 


40 


Epithelium, some nerve 




0.2 


8.1 


7.0 


19 


0.95 


8.8 


35 


Epithelium, little nerve 



Magnesium snlfate. Table IV summarizes the effects of various concentra- 
tions of magnesium sulfate applied for varying lengths of time. While all con- 
centrations used induce nerve if applied long enough, the higher concentrations 
induce spreading nerve while the lower concentrations induce radial nerve. 
Similarly, if the concentration is kept constant and the time varied, long exposures 
result in spreading nerve while with short exposures radial nerve develops. Thus, 

6.2 mg./ml. of magnesium sulfate induces some radial nerve at 1.3 hours; after 

4.3 hours all radial nerve is obtained, while after 5.5 hours the cultures are 
mostly spreading nerve. 

It is of interest to point out that the presumptive epidermis survives continuous 



INDUCTOR INTENSITY 



133 



culture in concentrations as high as 5.2 mg./nil. This is a 25-fold increase in 
the concentration of magnesium sulfate used in the standard culture medium. 

Ammonium snlfatc. The effects of ammonium sulfate at two concentrations 
are given in Table V. At the lower concentration, 0.024 mg./ml., the presumptive 
epidermis survives 21 hours' exposure, but not continuous treatment. Three 
hours' treatment has no visible action and the presumptive epidermis spreads 
out as a sheet of epithelial cells. After 5 hours some radial nerve is formed, 

TABLE IV 

The effects of magnesium sulfate (MgSOf7H<iO) on the differentiation 
of presumptive epidermis cells (stage 11) 



Treatment 


Culture 


Results 


Cone., 
mg./ml. 


Time, hrs. 


Temp., C. 


Cone., 
mg./ml. 


No. 
aggregates 


Types of cellular differentiation 


2.2 


8.3 


24 


0.2 


25 


Radial nerve 


2.2 


0.5 


24 


2.2 


25 


Radial nerve 


3.2 
4.2 
5.2 


4.0 

4.0 
0.25 


23 
23 

22 


3.2 
4.2 
5.2 


75 
75 
75 


Radial nerve, spreading nerve 
Spreading nerve, radial nerve 
Spreading nerve 


6.2 
6.2 


1.3 
4.3 


22 

22 


0.2 
0.2 


25 

25 


Epithelium, radial nerve 
Radial nerve 


6.2 
6.2 


5.5 
1.5 


22 
23 


0.2 
0.2 


25 
35 


Spreading nerve, radial nerve 
Radial nerve 


6.2 


4.5 


23 


0.2 


40 


Radial nerve 


8.2 


2.5 


24 


0.2 


35 


Radial nerve 


8.2 
8.2 
8.2 


6.0 
24.0 
28.0 


24 
23.5 
23.5 


0.2 
0.2 
0.2 


40 
35 
40 


Spreading nerve 
Spreading nerve 
Spreading nerve 


10.2 
10.2 
10.2 
10.2 


1.3 
4.3 
5.5 
1.5 


22 
22 
22 
22 


0.2 
0.2 
0.2 
0.2 


25 
25 
25 
40 


Spreading nerve, radial nerve 
Spreading nerve 
Spreading nerve 
Radial nerve 


10.2 


4.5 


23 


0.2 


35 


Spreading nerve 


15.2 


2.0 


22 


0.2 


35 


Radial nerve 


15.2 


3.0 


22 


0.2 


40 


Radial nerve 



and after 8 and 21 hours' treatment all cultures exhibit radial nerve. Continuous 
exposure to 0.024 mg./ml. results in death. 

An increase in tolerance to 0.024 mg./ml. of ammonium sulfate is shown by 
the results of transfers from standard solution to ammonium sulfate at various 
intervals. After 3 and 5 hours the presumptive epidermis begins to spread 
in 0.024 ammonium sulfate but soon dies, while after 8 and 21 hours epithelial 
sheets and ciliated masses form and persist in ammonium sulfate. 

At a concentration of 0.25 mg./ml. ammonium sulfate induces radial nerve 
after 3 hours' exposure time, nerve after 5 hours and causes death after 7.5 hours' 
exposure time. 



134 LESTER G. EARTH AND LUCENA J. EARTH 

TABLE V 

The effects of ammonium sulfate on differentiation of presumptive epidermis cells (stage 11) 



Treatment 


Culture 




Cone., 

mg./ml. 


Time, hrs. 


Temp., C. 


Cone., 
mg./ml. 


No. 
aggregates 


Results 


.024 


3 


20-21 





25 


Epithelium 


.024 


5 


20-21 





25 


Epithelium, radial nerve 


.024 


8 


20-21 





25 


Radial nerve 


.024 


21 


20-21 





30 


Radial nerve 


.024 


3 


20-21 


.024 


25 


Dead 





3 


20-21 


.024 


25 


Epithelium dead 





5 


20-21 


.024 


25 


Epithelium > dead 





8 


20-21 


.024 


25 


Epithelium, ciliated masses 





21 


20-21 


.024 


15 


Ciliated masses 


.25 


1.0 


24.5 





35 


Epithelium 


.25 


3.0 


24.5 





35 


Radial nerve 


.25 


5.0 


24.5 





40 


Nerve 


.25 


7.5 


24.5 





35 


Dead 



Calcium chloride. Various nerve patterns are obtained when calcium chloride 
is added to our standard salt solution. As Table VI shows, the presumptive 
epidermis in the standard salt solution with 0.04 mg./ml. of calcium chloride 
forms chiefly extensive sheets of epithelial cells with cilated patches. Only a few 
cells in the center of the sheet form nerve fibers. With 0.15 mg./ml. of added 
calcium chloride radial nerve is predominant, and when 0.25 mg./ml. is added the 
presumptive epidermis forms chiefly spreading nerve with few epithelial cells. 

Higher concentrations, 0.54 and 1.04, induce short nerve and a new nerve 
pattern which has been termed "Ca-induced nerve." This latter nerve pattern 
has some of the characteristics of "short nerve" and some of "spreading nerve." 

TABLE VI 

The effects of calcium chloride (CaCh-2H0) on the differentiation 

of presumptive epidermis cells (stage 11) 

Standard solution contains 0.04 mg./ml. 



Treatment 


Culture 


Results 


Cone., mg./ml. 


Time, hrs. 


Cone., mg./ml. 


No. aggregates 


0.04 
0.19 


2.0 
2.0 


.04 
.19 


75 
75 


Epithelium, little nerve 
Radial nerve 


0.29 
0.54 


2.0 

7.0 


.29 
.04 


75 
35 


Spreading nerve, radial nerve 
Short nerve 


0.54 


2.0 


.54 


35 


Ca-induced nerve 


0.54 


1.5 


.54 


75 


Ca-induced nerve 


1.04 


7.0 


0.04 


40 


Short nerve 


1.04 


2.0 


1.04 


40 


Ca-induced nerve 


2.54 


3.0 


0.04 


35 


Radial nerve 


2.54 


5.5 


0.04 


40 


Spreading nerve, short nerve 



INDUCTOR INTENSITY 



135 



There are very few epithelial cells in the periphery and many short nerve fibers 
but also some long nerve fibers. 

As with magnesium sulfate it is interesting to note that the cells tolerate 
continuously a 25-fold increase in the concentration of calcium chloride. 

Lithium chloride. The data on the effects of lithium chloride are too extensive 
to report in detail. Preliminary experiments indicated that the type of cellular 
differentiation could be controlled either by the concentration of lithium chloride 
or by the time of exposure. These correlations are seen in Table VII, where 
various concentrations have been applied to the presumptive epidermis for varying 
lengths of time. For example at 1.5 mg./ml. a 10-minute exposure results in 
epithelium with ciliated patches as in controls. As the length of treatment is 
increased we first obtain radial nerve, then spreading nerve and finally pigment 
ring cells. 

TABLE VII 

The effects of lithium chloride on the differentiation of presumptive epidermis cells (stage 11) 





Duration of treatment 


Cone., 




mg./ml. 


















10 min. 


15-30 min. 


1 hr. 


2 hr. 


4 hr. 


5 hr. 


6-9 hr. 


0.50 


Epithelium 




Epithelium, 


Epithelium 






Pigment 








nerve 


Radial nerve 
















Pigment 








0.75 






Nerve, 




Pigment, nerve 












epithelium 




Epithelium 






1.00 


Epithelium 


Epithelium 


Radial nerve 


Spreading nerve 


Pigment 


Pigment 












Radial nerve 








1.25 






Radial nerve 


Spreading nerve 


Pigment, nerve 


Pigment 




1.50 


Epithelium 


Epithelium 


Radial nerve 


Pigment 


Pigment, nerve 


Pigment 


Pigment 






Radial nerve 


Spreading nerve 


Spreading nerve 








2.00 


Epithelium 




Pigment, nerve 


Pigment 


Pigment 


Pigment 


Pigment 


2.50 










Pigment 






3.00 






Pigment, nerve 


Pigment 


Pigment 


Pigment 




3.50 




Nerve 


Pigment 
















Radial nerve 










4.00 




Radial nerve 


Pigment, nerve 


Pigment 


Pigment 






6.00 






Pigment, nerve 


Pigment, nerve 









Similarly if we use one-hour exposure then with a concentration of 0.5 mg./ml. 
we obtain mostly epithelium with ciliated patches as in controls. Increasing the 
concentration to 1.0 mg./ml. results in radial nerve. With 1.5 mg./ml. we 
begin to find spreading nerve, and with still higher concentrations we obtain 
pigment ring cells. 

Often two types of cellular differentiation are present in the same culture 
dish. This variability may be in part a result of the experimental procedure. 
The aggregates are prepared in lithium chloride and the time of exposure is 
measured from the time at which all the aggregates are prepared to the time of 
transfer to standard salt solution. Since it takes 20-25 minutes to prepare the 
150 aggregates usually used for an experiment, some aggregates will have been 
exposed for as much as 25 more minutes than others. Of course the excised 
explant from which the aggregates are prepared has been in lithium chloride for 
the 20-25 minutes required for preparation of the aggregates and this fact may 
possibly minimize the difference in times of exposure of the aggregates. 

During the course of the preceding experiments some of the aggregates were 
permitted to remain in lithium chloride continuously. While such treatment 



136 LESTER G. EARTH AND LUCENA J. EARTH 

usually resulted in early death of the cells, the lower concentrations, 0.56 mg./ml., 
sustained the cells for about three weeks. These cultures did not develop nerve 
nor pigment ring cells hut rather contained very large scattered cells which 
formed thick filaments containing granules (Figs. 4, 6). The filaments of adjacent 
cells made contact, forming a complex lattice work or network (Figs 3, 5). These 
cells are tentatively identified as neuroglia cells. 

Table VIII records the effects of continuous treatment of presumptive epidermis 
with concentrations of lithium chloride ranging from 0.10 mg./ml. to 2.00 mg./ml. 
The "neuroglia" type cell is obtained at concentrations of 0.56 and 0.65 mg./ml. 
A 7-hour exposure to these concentrations gives the expected results, namely, 
pigment ring cells and nerve as well as some epithelium. 

TABLE VIII 
The effects of lithium chloride on the differentiation of presumptive epidermis cells (stage 11) 

Cone. Treatment Results 

nig. /nil. 

0.10 continuous Epithelium with ciliated patches; mucus cells 

0.20 continuous Ciliated masses secreting large amount of mucus; some epithelial 

sheets with ciliated patches 

0.30 continuous Ciliated masses and some epithelial sheets ; mucus secreted 

0.38 continuous Ciliated masses with voluminous secretion of mucus containing 

pigment granules; astrocytes 

0.47 continuous Some ciliated masses with mucus; others form large scattered 

cells with protoplasmic filaments; neuroglia 

0.56 continuous All neuroglia 

7 hrs. Epithelium, scattered pigment ring cells, little nerve 

0.65 continuous All neuroglia 

7 hrs. Epithelium, pigment ring cells, nerve 

0.74 continuous Many loose cells, about 8-10 large neuroglia cells 

0.91 continuous Many loose cells, about 5-6 large neuroglia cells 

1.50 continuous Dead 

4 hrs. Pigment, nerve 

2.00 continuous Dead 

4 hrs. Pigment, nerve 

We next attempted to see if there were any other cell types between the small 
pigment ring cells obtained with 6-9 hours treatment with low concentrations of 
lithium and the "neuroglia" cells found during continuous treatment. 

Table IX reveals a new cell type resembling an astrocyte. These cells arise 
as large scattered cells similar to "neuroglia" but possess a large ring of pigment 
granules near the nucleus, as do the small pigment ring cells which give rise to 
melanophores. The "astrocytes," however, do not give rise to melanophores, and 
persist as star-shaped cells for about 4 weeks (Figs. 1, 2). 

The sequence of inductions following a time course is shown by the data for 
0.74 mg./ml. The epithelium is first induced to become a determined nerve cell. 
These determined nerve cells are then induced to become determined pigment cells. 
Further treatment induces the determined pigment cells to become determined 
"astrocytes" while continuous treatment induces the determined "astrocytes" to 



INDUCTOR INTENSITY 



137 



become "neuroglia" cells. These various cell types must be considered to be 
determined in the sense that placed in standard salt solution they self differentiate. 



DISCUSSION 



Induction bv various ions 



The induction of neural cells from the presumptive epidermis by various 
compounds is not a new phenomenon. Many substances of widely differing 
chemical constitution have been shown to be neural inductors. In some instances 
where the substances have been implanted the action has been assigned to a 
toxic or sub-cytolytic action. In other cases, where the presumptive epidermis 

TABLE IX 

The effects of lithium chloride on the differentiation of presumptive epidermis cells (stage 11) 
Cone., nig. /ml. Time, hrs. Results 



0.47 

0.56 
0.74 
0.74 

0.91 

0.91 
1.50 



19.5 

25.5 

continuous 

5.5 
20.5 

6.0 
21.0 

5.5 

28.0 

47.0 

continuous 

5.5 

28.0 

47.0 

continuous 

29.0 
46.5 
56.5 

29.0 
46.5 
56.5 



Pigment cells, epithelium 
Pigment cells, epithelium 
Neuroglia 

Epithelial cells, pigment cells 
Pigment cells 

Small pigment ring pigment cells 
Large pigment ring astrocytes 

Epithelium, few pigment cells, little nerve 

All pigment cells 

Large pigment ring > astrocytes 

Neuroglia 

Pigment cells, nerve, epithelium 
Large pigment ring > astrocytes 
Large pigment ring > astrocytes 
Neuroglia 

Pigment cells, astrocytes 

Astrocytes 

Astrocytes 

Many dead, a few astrocytes 
Many dead, a few astrocytes 
Many dead, a few astrocytes 



is immersed in a solution of the substances, as in a conditioned medium (Nitt 
and T witty, 1953), it is difficult to accept a toxic action since the cells survive 
continuous exposure. There is of course the possibility that some cells are 
killed by the added substances and that the induction is actually brought about 
by the dead or injured cells. This latter possibility is always present since there 
is no way of obtaining presumptive epidermis without some injury caused by 
manipulation or operation. All we can do is prepare adequate controls for injury. 
Our controls consisting of presumptive epidermis from the inner layer of the 
ectoderm of stage 1 1 gastrula now differentiate into epithelial sheets with ciliated 
patches. Only rarely does nerve appear. The injury phenomenon is definitely 
present and in our cultures there are always some unattached dead cells which 
appear about two days after the preparation of the aggregates. These dead cells 



LESTER G. EARTH AND LUCENA J. EARTH 

do not induce nerve, however, under the conditions of our experiment. Therefore 
it is probable that the induction by such ions as Mg ++ , Ca ++ and HCOs~ is a result 
of the alteration of the living cells and not a secondary phenomenon caused by 
the presence of dead cells. The cells survive continuous treatment with these ions. 
The significance of the experiments on induction with Mg ++ , Ca ++ and HCOs~ 
resides in the fact that they are normally present in living inductor cells and a 
release of any or all of these ions during gastrulation would result in induction 
of a neural plate. On the other hand, the normal inductor may bring about the 
release of any or all of these ions within the ectoderm cells with which it comes 
in contact. With these widely differing interpretations of the action of Mg ++ , 
Ca ++ and HCOs" possible we do not feel that these studies throw any light on the 
naturally occurring inductors in the chorda mesoderm. 

SEQUENTIAL INDUCTION BY LITHIUM CHLORIDE 

While the lithium ion can be definitely excluded as a natural inductor the 
sequence of inductions obtained by varying the concentration of lithium chloride 
may be of significance in the interpretation of the action of a naturally occurring 
inductor regardless of its chemical constitution. 

Let us use epi for epidermis; NI for radial nerve; N 2 for spreading nerve; 
P for small pigment ring cells ; A for "astrocytes" and Ng for "neuroglia" cells. 
Then we may formulate a hypothesis of 6 different pathways from presumptive 
epidermis to the 6 cell types : 

undifferentiated ectoderm cell 

I I I I I I 
epi Ni N 2 P A NG 

The various potencies of the presumptive epidermis are shown here as diverse 
pathways. Lithium chloride in low concentration of 1.25 mg./ml. for one hour 
might be supposed to inhibit all pathways but that leading to NI, i.e., radial nerve. 
The cells are thus determined and will differentiate into nerve cells in standard 
solution. If, however, these determined nerve cells be exposed to lithium chloride 
for four hours the cells will differentiate into pigment cells (P) in standard solu- 
tion. Thus, the pathway toward P was not destroyed by the one-hour treatment. 
Similarly, it can be shown that after P is determined, further treatment with 
lithium chloride results in A or Ng. Thus, lithium chloride does not appear to 
act by inhibiting all pathways but one, and the concept of diverse pathways is 
not applicable. Rather the pathway to P, for example, must pass through NI 
and No. 

We meet with a similar inconsistency if the assumption is made that lithium 
chloride in low concentration stimulates the pathway leading to NI while a high 
concentration stimulates the pathway to P. For again we would assume that the 
other pathways are lost when NI is determined. But further treatment after 
NI is determined results in No, still further in P, and with very long times A 
results. Thus, the pathways do not appear to be divergent but rather they are 
probably arranged in some linear order. 

The foregoing criticisms of the diverse-pathways concept leads to the formula- 



INDUCTOR INTENSITY 139 

tion of a sequence of potencies. The potencies would be present as a chain of 
reactions with epi and Ng at the ends of the chain. Whether we begin with 
epi or end with it depends upon whether we choose stimulation or inhibition as 
the mechanism for the induction. For example : if 

12345 

epi - Ni -> N 2 -> P -* A -> Ng, 

then Lii could stimulate at 1, higher concentrations or longer times would stimulate 
2, 3, 4, 5 progressively. Thus, at any time the cells might be determined to 
form P but the possibility of A and Ng still exists. In this formulation cells 
may be determined to form NI without losing the potencies in the rest of the 
sequence. 

On the other hand if we write the sequence 

54321 

Ng - A -> P -> N 2 -> Ni -> epi, 

then we express a concept of a series of reactions leading to the formation of 
epidermis unless stopped at some step. Mild inhibition by lithium chloride would 
then block the reaction 1, thus determining NI, while more inhibition would block 
at 1 and 2, resulting in N 2 . Still more inhibition might be expected to block 
1, 2, 3, etc. Such a formulation retains all the reactions determining P when 
1 and 2 are blocked, resulting in the determination of N 2 . There would probably 
be some time limit on the possibility of determining P after N 2 is determined 
since the formulation shows the reactions proceeding from left to right. Presum- 
ably, with steps 1 and 2 blocked all the products of the sequence of reactions would 
be used in the differentiation of No. However, early in the process N 2 may be 
determined so that it will self differentiate in absence of induction but P may still 
be determined by a longer treatment with lithium chloride. This latter statement 
is simply a restatement of the results of the experiments reported in this paper. 

Much of the research dealing with induction of the amphibian presumptive 
epidermis is difficult to interpret in light of the naturally occurring inductor or 
inductors. The present study is no exception. At first sight it is tempting 
to draw the parallel between the varying times of action of the organizer as 
gastrulation proceeds and the effects of lithium chloride at varying times of 
exposure to the presumptive epidermis. However, Mangold's transplantation of 
different regions of the roof of the archenteron to the blastocoele has shown that 
head induction and trunk induction may be obtained when the times of action 
are the same (Mangold, 1933). Thus, while it is definitely true that the 
posterior presumptive neural plate is exposed to the invaginating roof of the 
archenteron several hours before the anterior presumptive neural plate, this fact 
seems to be irrelevant to the induction process. The posterior neural plate may be 
induced by the posterior roof archenteron without previous contact by the anterior 
roof of archenteron. 

It is true, however, that in the process of gastrulation the anterior roof of 
the archenteron first induces the presumptive spinal cord to become determined 
as forebrain. This determined forebrain then becomes determined as hindbrain 
under the influence of hindbrain inductor, and finally the determined hindbrain 



140 LESTER G. EARTH AND LUCENA J. EARTH 

is induced to become spinal cord. Thus, the preliminary induction of forebrain 
does not exclude the later induction of spinal cord and in this respect normal 
induction resembles sequential induction by lithium chloride. 

It is also tempting to conclude that, since different concentrations of lithium 
chloride induced different cell types, the organizer is merely a graded series of 
concentrations of a single inductor substance. And indeed there is supporting 
evidence from the report by Yamada (1958) on the changes of inductive ability 
of the bone marrow upon heating. Possibly also the fact that denaturation of 
the kidney pentose nucleo protein inductor changes its inductive power from 
posterior neural induction to anterior neural induction might be cited (Yamada, 
1958). 

On the other hand, the experiments with dead roof of archenteron show no 
differences in inductive ability between anterior and posterior regions. If different 
'Concentrations of some inductor were responsible for anterior and posterior neural 
plate, one would expect these to be present in the dead cells. For the dead cells 
are able to induce although they lost the regional specificity for induction. The 
situation becomes complex indeed when we begin to suggest that the induction 
by dead archenteron roof is not brought about by the natural inductor. 

Perhaps the fairest statement would be since lithium chloride in various con- 
centrations induces various cell types, it makes possible an interpretation of the 
organizer as a concentration gradient of a single inductor, but the experiments 
presented in this paper do not constitute direct evidence. 

SUMMARY 

1. Presumptive epidermis of stage 11 gastrula differentiates into epithelial 
sheets with ciliated patches. 

2. Alteration of the basic salt solution by addition of calcium chloride or 
magnesium sulfate or sodium bicarbonate results in the differentiation of nerve cells. 

3. Calcium chloride, magnesium sulfate, sodium bicarbonate, ammonium sulfate 
and lithium chloride will induce nerve cells from presumptive epidermis when 
applied for a few hours beginning with stage 11. 

4. Lithium chloride applied to the presumptive epidermis for varying lengths 
of time induces various cell types. 

5. Lithium chloride applied in various concentrations to the presumptive 
epidermis induces various cell types. 

6. A concept of sequential induction is introduced as a formal explanation of 
the lithium chloride inductions. 

LITERATURE CITED 

EARTH, L. G., AND L. J. EARTH, 1962. Further investigations of the differentiation in vitro 

of presumptive epidermis cells of the Rana pipicns gastrula. /. Morphol., Ill : 347-373. 
MANGOLD, O., 1933. Uber die Induktionsfahigkeit der verschiedenen Bezirke der Neurula von 

Urodelen. Naturwiss., 43 : 761-766. 
Niu, M. C, AND V. C. TWITTY, 1953. The differentiation of gastrula ectoderm in medium 

conditioned by axial mesoderm. Proc. Nat. Acad. Sci., 39 : 985-989. 
SHUMWAY, WALDO, 1940. Stages in the normal development of Rana pipiens. Anat. Rec., 

78: 139-147. 
YAMADA, T., 1958. Embyronic Induction. In: The Chemical Basis Of Development; pp. 

217-218. Ed. by W. D. McElroy and B. Glass, Johns Hopkins Press, Baltimore, Md. 



VARIATIONS IN THE LARVAL STAGES OF A DECAPOD CRUSTA- 
CEAN, PLEURONCODES PLANIPES STIMPSON (GALATHEIDAE) 1 

CARL M. BOYD 2 AND MARTIN W. JOHNSON 

Scripps Institution of Oceanography, La Jolla, California 

During the last decade extensive literature has been accumulating on the larval 
development of several species of decapod Crustacea. Costlow, Bookhout and 
Monroe (1960) have reared several species of Brachyura. Rees (1959) has 
reared Emerita talpoida, Coffin (1958) successfully raised species of pagurids, 
and Broad (1957a and 1957b) raised two species of Macrura (Palaemonetes} . 
The Brachyura, with rare exceptions, all pass through a constant number of molts, 
though duration of time spent in the various larval stages is influenced by tem- 
perature and salinity. Broad (1957b) found that the number of molts passed 
through, and the duration of the larval stages, in Palaemonetes were influenced 
by type of food available to the larvae. Among the Anomura the picture is less 
clear. Johnson and Lewis (1942), in a study of larvae of Emerita analoga from 
the plankton, described the morphology of several discrete stages, and reported 
several specimens which appeared intermediate between stage III and stage IV; 
the authors assumed that the intermediate forms indicated variation in the molting 
sequence. When Rees reared the larvae of E. talpoida, he noted that some went 
through a stage that was deleted by others. The precise molting history of indi- 
vidual larvae in many of the studies has been obscured in laboratory studies by 
the practice of rearing several larvae in the same container. In plankton studies 
the observers are generally unable to discern what degree of morphological varia- 
tion occurs within the confines of the same larval stage. 

In 1960 one of us (Boyd) described and figured five larval stages of P. planipes, 
based on specimens taken from the plankton in neritic waters off southern Baja 
California. The five stages were morphologically discrete, and it was assumed 
that each stage was passed through in a single molt. Since then the larvae have 
been reared through the larval stages, the immature phase, and on to maturity, a 
total span of about a year. Laboratory data support the validity of the five stages 
described and add a great deal of information concerning the larval development. 

Pleuroncodes planipes is an anomuran galatheid crustacean, about 9 to 11 cm. 
long as adults, and resembling a small homarid lobster. These red crabs, as they 
are commonly called, exist both as pelagic and benthic animals. The crabs occur 
along the coasts of California and Baja California, over a range from 16 N. to 
37 N. and have a distribution which is typically neritic in the pelagic phase. 
Distribution, growth rates and notes on the biology of the adults and post-larval 
crabs are presented elsewhere (Boyd, 1963). 

1 Contribution from the Scripps Institution of Oceanography, University of California, 
San Diego. 

2 Present address : Institute of Oceanography, Dalhousie University, Halifax, Nova Scotia. 

141 



142 



CARL M. BOYD AND MARTIN W JOHNSON 



REPRODUCTION 

Adult Plcuroncodes planifics females carry their eggs attached to their pleopods 
in the same manner as brachyuran and other anomuran crabs. Specimens in the 
laboratory bred and laid eggs readily ; the eggs produced were carried from 6 to 22 
days (generally about 14 days). During that time the eggs changed from a golden 
color to a dark amber, as the eyes of the embryos developed. All the larvae hatched 
out as swimming Stage I zoeae, and once hatching began, all the eggs carried by 

TABLE I 

Numbers and ratios of post-larval male and female Pleuroncodes planipes caught in the 

monthly CalCOfI plankton samples, from December, 1958, to August, 1960. 

The number of these females which were carrying eggs is also given. 



Cruise 


Number 


& % males 


Number & 


% females 


Number & 
females 


% of those 
gravid 


Month 


5812 


13 


(54) 


12 


(46) 





(0) 


December 


5901 


293 


(55) 


241 


(45) 


27 


(11) 


January 


5902 


169 


(51) 


161 


(49) 


58 


(36) 


February 


5903 


16 


(40) 


24 


(60) 





(0) 


March 


5904 


186 


(58) 


133 


(42) 





(0) 


April 


5905 


111 


(51) 


106 


(49) 





(0) 


May 


5906 


89 


(52) 


83 


(48) 





(0) 


June 


5907 


193 


(51) 


186 


(49) 





(0) 


July 


5908 


400 


(60) 


291 


(40) 





(0) 


August 


5909 


46 


(52) 


42 


(48) 





(0) 


September 


5910 


140 


(55) 


113 


(45) 





(0) 


October 


5911 


3 


(100) 





(0) 





(0) 


November 


5912 





(0) 


2 


(100) 


1 


(50) 


December 


6001 


199 


(59) 


141 


(41) 


41 


(29) 


January 


6002 


168 


(61) 


106 


(39) 


59 


(56) 


February 


6003 


324 


(49) 


334 


(51) 


133 


(40) 


March 


6004 


413 


(55) 


341 


(45) 


2 


(-6) 


April 


6005 


180 


(46) 


208 


(54) 





(0) 


May 


6006 


71 


(52) 


66 


(48) 





(0) 


June 


6007 


106 


(57) 


79 


(43) 





(0) 


July 


6008 


87 


(51) 


84 


(49) 





(0) 


August 


Total 


3207 


2753 









a female hatched within about 12 hours. Females carry up to 3650 eggs (the 
largest number counted in the study) ; larger females tend to have the greater 
number of eggs. Females kept isolated from males occasionally produced eggs, 
but these were invariably sterile, and were sloughed off from the pleopods within 
a few days. Records of individual females in the laboratory indicate that each 
female usually had two, and rarely three, broods of eggs per season. Sexual ma- 
turity, as denoted by the ability of the females to produce eggs, was attained at 
a size of 14-15 mm. standard carapace length. Females of that size are about 12 
months old. 

In the laboratory the egg-bearing season lasted from November through April, 
with the peak in February. Table I shows that the egg-bearing season in the field 
followed a similar pattern. 



VARIATIONS IN STAGES OF DECAPOD LARVAE 143 

The numbers of males and females caught in the plankton tows of the Cali- 
fornia Cooperative Oceanic Fisheries Investigations between December, 1958, and 
August, 1960, differed significantly from 50/50 at the 0.01 level, as tested by the 
signed ranks test, two-tailed. The average percentage observed was 53.85% males 
and 46.15% females. The reasons for this departure from the 50/50 sex ratio 
are not known, but the departure may result from one or more of the following: 
(1 ) more males may be hatched than females, (2) females may have a lower sur- 
vival rate than males, or (3) the plankton nets may not sample males and females 
with equal effectiveness. 

METHODS 

The rearing techniques used in this study are similar to those described by 
Broad, Coffin, Costlow et al., and Rees. They involve the use of antibiotics to 
reduce the numbers of contaminating bacteria in the larval cultures, and also the 
use of Artemia nauplii for larval food. Freshly hatched larvae of P. planipes, taken 
from adult females kept in the laboratory, were pipetted into containers filled with 
sea water taken from the end of the pier of the Scripps Institution of Oceanog- 
raphy, La Jolla, California, where this work was clone. The water had been 
filtered through glass wool to remove detritus and larger animals that might prey 
upon the larvae. 

Experiment 1; started 19 February, 1960; duration 74 days. Two hundred 
freshly hatched larvae were placed, 10 per container, in 20 one-liter styrene plastic 
containers, each holding 500 ml. of water with antibiotics. The antibiotics used 
were (1) 50 mg. /liter streptomycin (trade name "Combistrep" by Pfizer; dihydro- 
streptomycin and streptomycin sulfate ; powder), and (2) 50 mg. /liter penicillin 
(penicillin G by Abbott, pill form, buffered with CaCO 3 , 928 units per mg., ground 
to a powder before use). This penicillin was used because earlier experiments 
indicated that the more readily available penicillin (a mix of 75% procaine peni- 
cillin and 25% penicillin G) was toxic to crab larvae. The containers were placed 
in trays of flowing sea water, which held them at temperatures ranging from 15 C. 
in February to 18 C. in June. Larvae were transferred to fresh sea water twice 
each week, and freshly hatched Artemia nauplii were added at that time. In the 
process of transferring the larvae into the fresh medium, each was drawn up into 
a 2-mm. bore glass pipette and examined through the pipette under a low power 
dissecting microscope to determine its stage. 

Experiment 2; started 1 March, 1960; duration 74 days. One hundred larvae, 
10 in each of 10 containers, were treated similarly to the larvae in Experiment 1. 
except that the penicillin was omitted ; streptomycin was used in concentrations of 
50 mg./liter; temperatures ranged from 15 to 18 C. 

Experiment 3; started 12 April, 1960; duration 108 days. Ninety-three larvae 
were kept individually in plastic containers in 50 ml. sea water ; 50 mg./liter 
streptomycin were added. Temperature, 16 to 19 C. Larvae were fed, trans- 
ferred into new medium, and examined as in the preceding experiments. The 
containers were checked daily for exuviae and these, if present, were removed and 
preserved in glycerine on slides. 



144 



CARL M. BOYD AND MARTIN W. JOHNSON 



RESULTS OF EXPERIMENTS 1, 2, AND 3 

The experiments proved that the larvae could be reared through all larval stages 
in the laboratory, and gave information concerning the total duration of the larval 
phase. The data are summarized in Table II. 

Each of the seven larvae to complete the larval phase in Experiment 1, and 
one of the 56 in Experiment 2 passed through a new stage, VI. This stage has 
never been seen in a search of hundreds of larval specimens from the plankton. 
It was similar in morphology to Stage V, but was larger and characterized by a 
tuft of plumose setae on each of the pleopods. It is probable that it was an artifact 
of treatment, and its presence was due either to the penicillin or to the CaCO 3 
buffer in the pills used, for other conditions were similar. As later experiments 
involving mixtures of non-buffered penicillin and streptomycin did not give Stage 
VI larvae, it is possible that the carbonate was the cause. 

It was noted in the first two experiments that Stage IV had a longer duration 
than the other stages. Modal values in Experiment 2 were: Stage I, 9 days; 
Stage II, 8 days; Stage III, 8 days; Stage IV, 25 days; Stage V, 12 days. The 
mean durations could not be calculated because larvae were followed as a popula- 

TABLE II 
Duration and survival of larvae in Experiments 1, 2 and 3 



Exp. 


Shortest duration of 
larval phase 


Longest duration of 
larval phase 


Average 


Survival through 
larval phase 


1 
2 
3 


54 days 
53 days 
71 days 


68 days 
74 days 
110 days 


61 days 
64 days 
87 days 


7/200 = 3.5% 
56/100 = 56% 
15/93 = 16% 



tion, and not as individuals. It was suspected that Stage IV consisted of at least 
two sub-stages, but the exact number of sub-stages or the morphological differences 
between them could not be determined because the history of individual specimens 
could not be followed. Experiment 3 was set up to make it possible to follow indi- 
vidual specimens through the larval stages. 

Although the volume of water per larva was identical (50 ml.) in each of the 
three experiments, and other conditions were the same in Experiments 2 and 3, the 
duration of the larval phase was longer in Experiment 3 than in either 1 or 2 
(cj. data presented above). The three average values cannot be compared in the 
usual manner because the duration of the larval period in Experiments 1 and 2 
is not known for individual larvae, and a variance cannot be calculated. A com- 
parison of the ranges, however, indicates that the values for Experiment 3 differ 
significantly (at better than the 0.05 level) from the values for Experiments 1 and 
2, but that the latter do not differ between themselves. The slightly greater tem- 
perature of Experiment 3, if it had any effect, should have caused a shortening of 
the larval period (see below). The only parameter in the three experiments which 
was known to be very different was the isolation of individual larvae in Experiment 
3 versus their group rearing in Experiments 1 and 2. This difference had two 
components : the larvae had less total water in which to swim, though the volume 



VARIATIONS IN STAGES OF DECAPOD LARVAE 



145 



TABLE III 

The percentage of larvae passing through each sub-stage of Stage I V 

(Experiment 3) 



IVa 
IVb 
IVc 

IVd 



100% 
100% 
100% 
100% 



IVe = 82% 
IVf = 65% 
IVg = 59% 
IVh = 29% 
IVi = 2% (died before reaching Stage V) 



of water per larva was identical, and they were not in association with other larvae. 
Either may have prolonged the larval phase in Experiment 3. 

By following the molting sequence of individual larvae it became evident that 
the number of molts passed through before a larva completed the larval phase 
varied between larvae. The morphology of Stages I, II, and III in the laboratory 
was as described in 1960, based on larvae taken from plankton collections. Each 
of these stages was completed after a single molt. Stage IV, however, was divided 
into a series of what can be called sub-stages, each separated by a molt, but all 
morphologically within the general description of Stage IV. The larva increased 
in size through the sequence of sub-stages, and the number of setae on some of 
the appendages (e.g., uropods, antennal scales) increased, but the differences be- 
tween the various sub-stages of the complex were so inconsistent that a sub-stage 
could be identified with certainty only by knowing how many molts the larva had 
passed through. The number of sub-stages within the Stage IV complex varied 
from four to nine. The percentage of larvae passing through each sub-stage is 
given in Table III. 

The mean duration of each of the larval stages and the 95% confidence limits 
of the sample, the cumulative elapsed time to the end of each particular stage (2), 
the number of larvae completing each stage (N), and the instantaneous death rate 
based on N t N e~ dt are given in Table IV. 

TABLE IV 

The mean duration of each larval phase and the 95% confidence limits of the sample. 

The cumulative elapsed time to the end of each particular stage (2), and the 

number of larvae completing each stage (N) 



Stage 


Mean duration and 
95% limits 
(days) 


s 

(days) 


AT 


Instantaneous 
death rate 
(days) 


I 


11.9 5.7 


11.9 


46 


0.059 


II 


8.0 4.5 


19.0 


45 


0.003 


III 


7.4 5.7 


27.3 


41 


0.013 


IVa 


6.9 3.8 


34.2 


39 


0.007 


IVb 


6.7 5.9 


40.9 


36 


0.012 


IVc 


7.4 4.3 


48.3 


32 


0.016 


IVd 


8.8 5.4 


57.1 


28 


0.015. 


IVe 


8.7 4.4 


65.8 


22 




IVf 
IVg 
IVh 


11.1 6.6 
9.8 4.7 
10.0 4.0 


76.9 
86.7 
96.7 


18 
16 

8 


0.016 


V 


13.7 6.6 


86.9 


15 


0.000 



146 



CARL M. BOYD AND MARTIN W. JOHNSON 



The instantaneous death rate cannot be tabulated for individual stages subse- 
quent to sub-stage IVd, for the deletion of stages is confused with mortality in the 
data analysis. It is evident from the table that the highest death rate occurred in 
the first stage, and mortality after that was essentially equal one stage to the next. 



opproxima te ti me of 
f i rst larval molt 




1234 56 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 

Time in days; percent mortality of larvae of 
Pleuroncodes p/anipes reared at several constant 
temperatures. 

FIGURE 1. Mortality of larvae of Pleuroncodes planipes reared in the laboratory at several 

constant temperatures. 

EXPERIMENT 4 
METHODS 

Tliis experiment started April 25, 1960; duration 163 days. A device was de- 
signed and constructed which maintained larval cultures at six constant tempera- 
tures. The temperatures selected were 12, 14, 16, 18, 20, and 22 C. The 
two extreme temperatures selected are approximately the surface temperatures of, 
respectively, the northern and southern ends of the distributional range of the 
.adults in the spring months (the breeding season). Larvae were placed in sea 
water containing 50 mg./liter penicillin (Pfizer, penicillin G, potassium; 1585 
tmits/mg. powder) and 50 mg./liter streptomycin (Combistrep, by Pfizer). Larvae 
were transferred into fresh sea water twice each week, and at that time were staged 
under the microscope and fed as in the previous experiments. The larvae were 
kept in styrene containers, each containing 18 compartments which measured 4.5 X 
5.0 X 3.8 cm. deep and held 50 ml. of sea water. Six containers were used at 



VARIATIONS IN STAGES OF DECAPOD LARVAE 



147 



each temperature; initially, two larvae were placed in each compartment, giving 
216 larvae at each temperature, or a total of 1296 larvae. All of these larvae were 
obtained from the same female over a period of about 12 hours. The extra larva 
was placed in each compartment because Experiment 3 had shown that mortality 
was highest in the first few days. After 22 days the extra larvae in the 12, 14. 
and 16 cultures were discarded, leaving individual larvae (108 total) in the com- 
partments ; mortality had been higher at the higher temperatures, so that only 
101 larvae remained at 18, 16 larvae at 20 ; and none at 22. 



v 

IVh . 
IVg 
IVf 
IVe 



la rva I 
stage 



IVd . 
IVc . 
I Vb . 
IVa . 




10 20 30 40 50 60 70 80 90 100 110 120 130 

Mean number of days to the end of each larval stage of 
Pleuroncodes planipes reared at several constant temperatures 

FIGURE 2. Rates of development of larvae of Pleuroncodes planipes reared in the laboratory 

at several constant temperatures. 



RESULTS OF EXPERIMENT 4 

The mortality data for the first 22 days are shown in Figure 1. The estimated 
instantaneous death rates through Stage 1 (0-11.9 days) are: 12, 0.011; 14, 
0.018; 16, 0.020; 18, 0.020; 20, 0.074; 22, 0.192. No larvae completed the 
larval phase in either the 20 or 22 cultures; these temperatures may therefore 
be regarded as lethal in this experiment. 

Within certain limits larval development should be faster at high temperatures ; 
Figure 2 shows that this is true for P. planipes, at least over the range from 12 
to 18 C. No data are available for the duration of Stage I because at the time 
the larvae were in that stage they were not kept individually. In Figure 2 it will 
be noted that the cumulative mean number of days to the end of sub-stage IVh 



148 



CARL M. BOYD AND MARTIN W. JOHNSON 



is greater than the developmental time to the end of Stage V for the 14 and 16 
cultures. This is because many larvae omitted one or more of the later Stage IV 
sub-stages and passed directly to Stage V, thereby shortening their larval duration. 
The trend of the 20 line is probably correct but it is based on too few individuals 



TABLE V 

Mean duration of stages of P. planipes larvae reared at several temperatures. These 

durations are cumulatively summed and presented under the heading 2. The 

number of larvae passing through each stage is presented under N. 



12 


14 


Stage 


Mean duration and 
95% limits 


V 


N 


Stage 


Mean duration and 
95% limits 


2 


N 


II 




24.7 




II 




22.3 




III 


16.9 7.0 


41.6 


84 


III 


13.2 8.1 


35.5 


61 


IVa 


12.8 5.4 


54.4 


71 


IVa 


12.6 9.6 


48.1 


52 


IVb 


10.7 5.8 


65.1 


68 


IVb 


9.9 5.9 


58.0 


45 


IVc 


10.4 6.7 


75.8 


63 


IVc 


9.5 6.8 


67.5 


42 


IVd 


9.1 5.1 


84.6 


60 


IVd 


9.3 5.2 


76.8 


37 


IVe 


11.7 5.1 


96.3 


56 


IVe 


7.2 4.0 


84.0 


35 


IVf 


12.4 7.0 


108.7 


50 


IVf 


9.1 3.8 


93.1 


33 


IVg 
IVh 


12.0 6.7 


120.7 


13 


IVg 
IVh 


8.9 3.0 
9.0 5.7 


102.0 
111.0 


15 
2 


V 


17.7 7.9 


127.5 


50 


V 


13.2 4.7 


110.9 


31 


16 


18 


Stage 


Mean duration and 
95% limits 


V 


N 


Stage 


Mean duration and 
95% limits 


2 


AT 


II 




19.7 




II 




19.0 




III 


9.6 7.1 


29.3 


57 


III 


7.8 6.0 


26.8 


33 


IVa 


12.8 6.6 


42.1 


51 


IVa 


10.6 6.5 


37.4 


12 


IVb 


10.0 6.9 


52.1 


43 


IVb 


8.8 4.8 


46.8 


9 


IVc 


8.8 7.0 


60.9 


37 


IVc 


9.2 4.2 


55.4 


5 


IVd 


8.1 5.8 


69.0 


34 


IVd 


7.6 2.8 


63.0 


5 


IVe 


7.6 4.7 


76.6 


32 


IVe 


9.0 3.6 


72.0 


4 


IVf 


7.4 5.9 


84.0 


28 


IVf 


6.8 1.1 


78.8 


4 


IVg 
IVh 


7.2 4.3 
8.5 4.3 


91.2 
99.7 


17 

2 


IVg 
IVh 


4.0 
6.0 


82.8 
88.8 


2 

1 


V 


10.6 3.3 


98.4 


30 


V 


11.7 1.7 


97.7 


3 


20 




Stage 


Mean duration and 
95% limits 


s 


N 


II 




18.0 




III 


7.0 


25.0 


3 




IVa 


14.5 3.5 


39.5 


2 




IVb 


10.0 


49.5 


1 




IVc 


4.0 


53.5 


1 




IVd 


7.0 


60.5 


1 





VARIATIONS IN STAGES OF DECAPOD LARVAE 



149 



TABLE VI 

A. The number of larvae which molted directly to Stage V from each sub-stage, 

thereby omitting later sub-stages 

B. The above data are expressed as the per cent of larvae which completed a 

particular sub-stage before becoming Stage V larvae 





12 


14 


16 


18 


IVe 


3 


1 


3 





IVf 


37 


17 


11 


1 


IVg 


13 


13 


15 


1 


IVh 





2 


2 


1 


V 


53 


33 


31 


3 


IVe 


100 


100 


100 


100 


IVf 


94 


97 


90 


100 


IVg 


24 


45 


55 


67 


IVh 





6 


6 


33 



to be very accurate; at the end of Stage III only three larvae were alive in the 
20 culture. The mean duration of each stage (with the 95% confidence limits of 
the sample) is shown in Table V. The values are cumulatively summed to give 
the average total number of days elapsed to the end of each stage. 

A O 10 value for the rate of larval development of P. planipes can be calculated 
on the basis of the mean number of days of life to the end of the larval phase. 
The first three values, 128, 111, and 98 (for 12, 14, and 16, respectively), give 
an average Q 10 of 1.95. The fourth value, for 18 (98), when compared with 
the 12 value gives a Q 10 of 1.6. This figure is based on only three larvae at 18, 
and is suspect; the correct value is probably about 1.9. 

The number of sub-stages passed through in Stage IV varied from larva to 
larva in this experiment as it did in Experiment 3. 

From Table V it can be seen that no larvae in the 12 culture went into sub- 
stage IVh ; at other temperatures some larvae went through this sub-stage on their 
way to Stage V. The number of larvae which molted directly to Stage V from 
each sub-stage, and thereby omitted later sub-stages, is tabulated in Table VI (A). 
These data may be expressed as the per cent of larvae which completed a par- 
ticular sub-stage before becoming Stage V larvae (Table VI (B)). A Friedman 
analysis of variance can be applied to the last three rows of figures in Table VI (B). 
This tests whether the columns of numbers come from the same population, and 
yields, in this case, a probability of 0.075. If one regards this value as significant, 
the data are evidence of a trend indicating that at higher temperatures larvae pass 
through more sub-stages than they do at lower temperatures. This is in spite of 
the fact that the total larval span in time is shorter at higher temperatures. 

DISCUSSION 



Perhaps the most interesting observation to result from the culturing of larvae 
of P. planipes in the laboratory is that Stage IV is divided into sub-stages, and 



150 CARL M. BOYD AND MARTIN W. JOHNSON 

that the number of these sub-stages may vary. The data indicate that the number 
of sub-stages which a larva passes through may be influenced by the temperature 
of the environment, with higher temperatures producing more sub-stages before 
the molt to Stage V ; that within limits the rate of larval development shows a Q 10 
close to 2 ; and that higher temperatures cause a higher death rate. 

Environmental factors other than temperature affect the duration of the larval 
stages. For example, temperatures were the same in Experiments 2 and 3, but 
there was a difference in the mean larval duration. The only known differences in 
conditions were the size of the rearing container and the presence of other larvae 
in the same container. Gurney (1942) conjectured that data concerning the life- 
histories of laboratory-reared larvae might prove misleading when applied to larvae 
in the ocean. In view of the demonstrated effects of various small changes of con- 
ditions on the molting sequence and developmental duration of laboratory-reared 
larvae, it is quite possible that their life-histories may be inapplicable to larvae 
in the field. The matter is further discussed by Rees (1959). In the case of P. 
planipcs larvae, however, all of the stages, except Stage VI, seen in the laboratory, 
including evidence for a complex of Stage IV sub-stages, can be found in larvae 
from the plankton. The number and detailed morphology of the Stage IV sub- 
stages which larvae pass through in the ocean may well differ from the number 
passed through by larvae in laboratory experiments, and the number of sub-stages 
may even vary from one part of the ocean or one season of the year to another. 

The irregularities in the number of molts in the larval phase shown by P. planipes 
may be found in other anomuran crabs. Sub-stages, however, are difficult to detect 
in morphological studies of larvae taken from plankton collections. Johnson and 
Lewis (1942) found a "lower Stage IV" in the larval stages of Emcrita analoga, 
based on larvae from the plankton. This lower stage is best interpreted as a sub- 
stage and is an indication that sub-stages occur naturally in the field. Rees ( 1959 ) 
noted that larvae of Emerita talpoida reared in the laboratory may pass through 
either 6 or 7 molts before becoming megalops. Similar results were noted in 
laboratory-reared larvae of E. analoga by Dr. Ian Efford (unpublished results, 
personal communication). A. Provenzano (personal communication) has observed 
a varying number of molts in larvae of some pagurid crabs from Florida cultured 
by him. Costlow (personal communication), however, has observed variation in 
the number of larval molts in only two species of Brachyura out of a total of about 
20 species both Portunidae; variation occurred only occasionally and resulted in 
a form with reduced viability which only rarely developed to the megalops stage. 
Broad (1957a, 1957b) found that the number of larval molts varied in Palaemonetes 
pugio, a decapod macruran, depending on the type of food given the larvae. 

Stage VI, which appeared in Experiment 1, has never been found in the plankton 
and appears to be the result of laboratory rearing conditions. It is possible that it 
occurs in nature under certain conditions. Its existence would certainly support 
Gurney's contention that laboratory conditions produce aberrant larval forms. 
Kurata (1960) observed what may be a similar phenomenon in the advanced stages 
of two lithodid crabs (Paralithodes camtschatica and P. brevipes} reared in the 
laboratory. These stages were intermediate between the usual last larval stage 
and the glaucothoe. 



VARIATIONS IN STAGES OF DECAPOD LARVAE 151 

The temperatures in Experiment 4 (12 to 22 C.) were selected because these 
are approximately the surface temperatures at the northern and southern ends of 
the crab's distribution in the adult phase. The few data available (unpublished) 
indicate that the larval distribution is similar to the adult distribution, with the 
greatest concentrations occurring along the western coast of southern Baja Cali- 
fornia. In that area winter temperatures may commonly be as high as 20 at 
the surface. Larvae at that temperature in the laboratory had a higher mortality 
rate than did those at lower temperatures. Possibly, larvae in the latitude of 
southern Baja California do not live at the surface but rather below the surface 
at a more optimal temperature. The winter breeding season may be correlated 
with higher larval survival in the laboratory at colder temperatures. 

Because problems encountered by the larvae of many polychaetes and the deca- 
pod crustaceans are similar, in that they must transform from a pelagic phase to 
a benthic phase, it is tempting to speculate that decapod crustacean larvae such 
as those of P. planipes may respond to environmental parameters in a way similar 
to that demonstrated by Wilson for various polychaete larvae (cf. Wilson, 1952). 
He has shown that polychaete larvae are influenced by the nature of certain sub- 
strates so that they end the pelagic phase and become benthic. The presence of 
other substrates tended to prolong the larval phase. Experiments similar to 
Wilson's have not yet been performed on larval decapods, however, presumably 
because of the difficulties inherent in rearing them. 



The authors wish to acknowledge the help given by Mrs. Dorothy Walton and 
Mr. John Nordback, who acted as technical assistants during parts of the study. 
Dr. Maurice Blackburn provided financial support through Scripps Tuna Ocea- 
nography Research Program for some of the work, and his help is gratefully 
acknowledged. 

SUMMARY 

1. The young of Pleuroncodcs planipes pass through a series of five morpho- 
logically discrete zoeal larval stages after hatching, and, except in Stage IV, the 
larvae change from one stage to the next by a single molt. Larvae in Stage IV 
may molt from four to eight times without greatly altering their basic morphology. 
There is evidence from laboratory culturing that the number of these sub-stages 
may be influenced by the temperature at which the larvae develop, with higher 
temperatures causing more sub-stages. 

2. The duration of the larval phase is influenced by the temperature at which 
the larvae are reared, and the rate of development follows a Q ]0 of about 1.9. 
The larval duration is also influenced by rearing conditions other than temperature, 
for the size of the larval rearing container or the presence of other larvae in the 
container has also been shown to influence the duration of the larval phase. 

3. A larval stage was seen in the laboratory which has not been found in nature, 
and it is probable that the stage was an artifact of laboratory culturing conditions. 

4. It may be generalized that variation in the number of larval molts is wide- 
spread in the Anomura ; variation does not commonly occur in brachyuran Crustacea.. 



152 CARL M. BOYD AND MARTIN W. JOHNSON 

LITERATURE CITED 

BOYD, C. M., 1960. The larval stages of Plcuroncodes planipes Stimpson (Crustacea, Decapoda, 
Galatheidae). Blol. Bull, 118: 17-30. 

BOYD, C. M., 1963. Growth rates, distribution and notes on the general biology of a marine 
decapod crustacean, Plcuroncodes planipes Stimpson (Galatheidae). U. S. Fish and 
Wildlife Service Fishery Bulletin; in press. 

BROAD, A. C., 1957a. Larval development of Palaemonetes pugio Holthuis. Biol. Bull., 112: 
144-161. 

BROAD, A. C., 1957b. The relationship between diet and larval development of Palaemonetes. 
Biol. Bull., 112: 162-170. 

COSTLOW, J. D., JR., C. G. BOOKHOUT AND R. MONROE, 1960. The effect of salinity and tem- 
perature on larval development of Sesarma cinereum (Bosc) reared in the laboratory. 
Biol. Bull., 118: 183-202. 

COFFIN, H. G., 1958. The laboratory culture of Pagunts samuelis (Stimpson) (Crustacea, 
Decapoda). Walla Walla College Publications, Dept. Biol. Sciences and Biol. Station, 
No. 22 : 1-5. 

GURNEY, ROBERT, 1942. Larvae of decapod Crustacea. Ray Society, London. 306 pp. 

JOHNSON, M. W., AND W. M. LEWIS, 1942. Pelagic larval stages of the sand crabs Emerita 
analoga (Stimpson), Blepharipoda occidentalis Randall, and Lepidopa myops Stimp- 
son. Biol. Bull., 83 : 67-87. 

KURATA, H., 1960. Last stage zoea of Paralithodes with intermediate form between normal 
last stage zoea and glaucothoe. Bull. Hokkaido Reg. Fish. Res. Lab., 22 : 49-56. 

REES, G. H., 1959. Larval development of the sand crab Emerita talpoida (Say) in the labora- 
tory. Blol Bull, 117: 356-370. 

WILSON, D. P., 1952. The influence of the nature of the substratum on the metamorphosis of 
the larvae of marine animals, especially the larvae of Ophelia bicornis Savigny. 
Ann. Inst. Oceanographique, 27: 49-156. 






HIGH TEMPERATURE TOLERANCE AND THYROID ACTIVITY 
IN THE TELEOST FISH, TANICHTHYS ALBONUBES 1 

J. CHARLES CHEVERIE AND W. GARDNER LYNN 

Department of Biology, The Catholic University of America, Washington, D. C. 

Upper lethal temperatures have been ascertained for many species of fish 
(Loeb and Wasteneys, 1912; Hathaway, 1927; Sumner and Doudoroff, 1938; 
Brett, 1941, 1944, 1946, 1952; Doudoroff, 1942; Fry et al, 1942, 1946). Dif- 
ferent species, when tested under comparable conditions, exhibit characteristic 
diverse high-temperature death points which are often clearly related to the 
conditions of life to which the species are adapted. However, the precise mecha- 
nism by which death is caused at high temperature in ectotherms is still not 
understood (Fisher, 1958; Precht, 1958). The experiments of Fortune (1955) 
led to the conclusion that increased activity of the thyroid gland with increasing 
environmental temperature may be an important factor in the thermal death of 
fishes. Fortune reported that specimens of Phoxinus {phoxinus} laevis kept in 
.05% thiourea solution at 23 C. for three days and then subjected to a 10 
temperature rise over a two-day period were able to survive this treatment and 
to live indefinitely at 33 C. On the other hand, specimens given no thiourea 
treatment during the three-day period of acclimation at 23 C. all succumbed by 
the time the temperature reached 24 C. Inhibition of thyroid activity by thiourea 
treatment thus enabled Phoxinus to tolerate a temperature 9 C. above that which 
would otherwise have been lethal. Fortune also reported that the thermal range of 
Lebistes reticulatus could be similarly extended by thiourea treatment. 

Earlier work by Evropeitzeva (1949) on thiourea-treated fry of Coregonus 
is in agreement with Fortune's findings but several other investigations have given 
diametrically opposed results. LaRoche and Leblond (1954) found that the parr 
of Salmo salar thyroidectomized with radioiodine had a lower tolerance for high 
temperature than did untreated controls. Auerbach (1957), working with Lebistes 
reticulatus, Platys variatus, and Leuciscus rutilus, and Theobald (1959) with 
Gambusia affinis found that thiourea treatment decreased the high-temperature 
tolerance in these species. Suhrmann (1955) reported that the goldfish, Carassius 
vulgaris, has an increased high-temperature tolerance if kept in thiourea at a low 
acclimation temperature but has a decreased tolerance if the acclimation tempera- 
ture is high. 

This paper reports the results of a study of the effects of thiourea treatment 
upon high-temperature tolerance and upon thyroid histology in the cyprinid, 
Tanichthys albonubes. Effects upon pituitary histology, which were also investi- 
gated, will be considered in a separate publication. 

1 Supported by funds from United State Public Health Service Grant No. A-2921. 

153 



154 J. CHARLES CHEVERIE AND W. GARDNER LYNN 

MATERIALS AND METHODS 

Tanichthys albonubes Lin, the common whitecloud, is a small tropical ovipa- 
rous fresh-water fish. A total of about 300 specimens was used in this study. 
The average length of the fish used, measured from the tip of the snout to the 
hypural plate, was 2.6 cm. 0.2 cm. 

During the phases of the experiments described below, the control whiteclouds 
were kept in dechlorinated tap water and the experimental fish were kept in 
0.05% thiourea contained in either 5-gallon aquaria or polyethylene pails. These 
media received continual aeration and were changed twice a week. The fish were 
fed daily with dry tropical fish food. Prior to the actual experiments several 
series of pilot experiments were carried out before adopting the subsequent 
methods regarding the concentration and period of treatment with thiourea, the 
acclimation temperature to be used, and the method of determining the upper 
lethal temperature. 

1. Pre-acclimation period 

The pre-acclimation period was designed to bring about hypofunction of the 
thyroids of the experimental fish. To accomplish this, no more than 75 experi- 
mental fish were placed in a 5-gallon aquarium containing 18 liters of 0.05% 
thiourea and kept at room temperature (21 C.). An equal number of controls 
were also kept at room temperature in an equal volume of water in a similar 
aquarium. This phase was continued for 30^1-0 days. At the end of this period, 
histological examination of the thyroids of a few treated animals revealed that 
thyroid hypofunction had been effected. 

2. Acclimation at low temperature 

After the pre-acclimation treatment, the control and experimental whiteclouds 
were kept in 11 -quart polyethylene buckets containing thiourea solution or 
water. Each bucket contained no more than 40 fish. The buckets were placed 
in a 50-gallon Aminco Laboratory Bath (refrigerator type) regulated to maintain 
a temperature of 15.0 C. 0.05 C., and kept there from 40 to 60 days. During 
the pre-acclimation and acclimation phases, fish mortality was negligible. 

3. Upper lethal temperature determination 

A common method for determining the upper lethal temperature of fish is to 
directly transfer the test animals from the acclimation tank to the high-temperature 
tank and then to calculate the temperature at which half of the animals die in a 
given period of time. The periods of exposure to the high temperature vary 
widely in different studies. A number of investigators have used a 24-hour 
period (Black, 1953; Theobald, 1959). Following these authors, the lethal 
temperature in this study was considered to be that temperature at which 50% 
of the fish succumbed and 50% survived after 24 hours' exposure. High tempera- 
tures at which the whiteclouds were tested ranged from 29.5 C. to 31.5 C. The 
water baths used to test the fish at high temperatures were Aminco Laboratory 
Baths which could maintain water temperatures within 0.05 C. 



THYROID AND TEMPERATURE TOLERANCE 155 

It was observed in preliminary experiments that when whiteclouds were directly 
transferred from the acclimation tank to a high-temperature tank, symptoms of 
shock, such as rapid movement of the opercula, quick spiral swimming, and 
partial or total loss of equilibrium, occurred. To minimize this shock the method 
was modified as follows. Water baths were set at the high temperatures at 
which fish were to be tested. Samples of control and experimental fish (4-6 fish 
per sample) from the acclimation bath were placed in small polyethylene pails 
containing either water or 0.05% thiourea already cooled to 15.0 C. The pails 
were then transferred to a hot- water bath set at 45.0 C., and left there until the 
temperatures at which the fish were to be tested (29.5 C. to 31.5 C.) 
were reached. This required 15 0.5 minutes. The pails were then placed 
in the appropriate water baths set at the temperature at which that particular group 
of fish was to be tested. The fish were left at this temperature for 24 hours. 
This substitution of a method of increasing the temperature over a period of 15 
minutes for the direct transfer method originally used eliminated all signs of shock. 

4. Methods of fixation and staining 

After 24 hours' exposure to the upper lethal temperature, surviving control 
and experimental fish and an identical number of specimens from corresponding 
solutions in the 15.0 C. acclimation bath were removed and anesthetized in 
1/1000 tricaine-methane-sulfonate (MS-222). The MS-222 was warmed or 
cooled to the temperature from which the fish were taken. The lower jaws of 
these fish were fixed in Bouin's fluid and embedded in paraffin. Records were 
kept concerning the sex and maturity of all specimens. Serial sections of the 
thyroid glands from control and experimental fish from the same sex group and 
from all three phases of the experiment were stained simultaneously. The thyroids 
were stained in Gomori's chrome-alum-hematoxylin-phloxine. 

OBSERVATIONS 
1. Thyroid histology 

The thyroid histology of the experimental fish was significantly different from 
that of the control animals. The thyroids of control fish at the end of the various 
phases of the experiment showed no significant differences. Similarly, no sig- 
nificant differences were observed in the thyroid histology of the thiourea-treated 
fish after each of three experimental phases. The descriptions which follow are 
based on histological observations of the thyroids of control and experimental 
fish killed at the end of the acclimation phase of the experiment. 

Examination of serial sections of individuals from the control groups showed 
that the thyroid of the whitecloud belongs to the non-encapsulated diffuse type 
found in most teleosts (Fig. 1). The follicles are scattered in the connective 
tissue beneath the floor of the pharynx but tend to accumulate around the ventral 
aorta and the aortic arches. These follicles are roughly spherical in shape. The 
average number of follicles found in a representative cross-section through the 
thyroid region of control fish is 9. Occasionally small capillaries are seen in 
contact with some of the follicles. The follicular epithelium is mostly squamous, 
although some of the small follicles have cuboidal to low columnar epithelium, 



156 



J. CHARLES CHEVERIE AND W. GARDNER LYNN 



the average height being 3.0 microns (Fig. 2). The squamous cells contain 
little cytoplasm, and their flattened, dark-staining nuclei occupy most of the cell. 
The cuboidal epithelial cells have round, usually basally located, nuclei and a 
small amount of slightly basophilic cytoplasm. No colloid droplets or vacuoles 






i 

V*v #& f *lfc*> *'* :' 




FIGURE 1. Section through thyroid region of an untreated control kept at room tempera- 
ture (21 C.) for 30 days, then at 15 C. for an acclimation period of 60 days. SO X. 

FIGURE 2. A single thyroid follicle from the specimen shown in Figure 1. 485 X. 

FIGURE 3. Section through thyroid region of a fish kept in 0.05% thiourea throughout a 
preacclimation period of 30 days at 21 C. and an acclimation period of 60 days at 15 C. 50 X. 

FIGURE 4. A single thyroid follicle from the specimen shown in Figure 3. 485 X. 

were observed in any of the follicular cells of the control thyroids. Except for 
a few of the small, more active follicles, the follicular lumina are large and filled 
with acidophilic, homogeneous, or granulated, non-vacuolated colloid. 

Examination of the thyroids of thiourea-treated fish reveals a histological 
picture quite different from that of the controls (Fig. 3). A great increase in the 
number and size of follicles has occurred so that the follicles are more compactly 
arranged and occupy a large portion of the sub-pharyngeal region. The average 
number of follicles found in a representative cross-section through the thyroid 
region of experimental fish is 25. The follicular epithelium is columnar and 



THYROID AND TEMPERATURE TOLERANCE 



157 



averages 19.5 microns in height. The cytoplasm of the thyroid epithelium con- 
tains large vacuoles and intracellular colloid droplets not found in controls (Fig. 
4). Large, round, vesicular, basally located nuclei containing prominent nucleoli 
are found in most of the follicular cells. Many follicles are collapsed while those 
which are not contain only a small amount of basophilic, vacuolated colloid. A 
hyperemic state is evidenced by the increase in the number and size of inter- 
follicular capillaries and other small blood vessels when compared with the controls. 

TABLE I 

Survival of fish previously acclimated at 15 C. and then exposed for 24 hours to the 

high temperatures indicated 



Test 


29.5 


30.0 


30.5 


31.0 


31.5 




#1 






5/6 


6/6 


0/3 




#2 








3/6 


0/6 


en 


#3 






3/6 


3/5 




8 

M 


#4 








3/6 




o 


#5 








2/6 




U 


#6 








3/6 






#7 








0/6 






#8 








6/6 




Totals 






8/12 
(67%) 


26/47 

(55%) 


0/9 




#1 


3/6 


3/4 


0/4 


0/4 






#2 




4/6 


2/6 






1 

Sn ^ 


#3 




3/6 


2/6 






S-2 


#4 




3/6 


1/5 






o g 


#5 




4/5 


0/5 






g* 


#6 




3/6 










#7 




1/6 










#8 




3/6 








Totals 


3/6 

(50%) 


24/45 
(53%) 


5/26 
(19%) 


0/4 





2. Upper lethal temperature 

Based on the method described above, the upper lethal temperature for control 
Tanichthys albonubes was found to be 31.0 C., while that for the experimental 
fish was 30.0 C. The results of the experiment are summarized in the accompa- 
nying table. The upper half of the table contains the data obtained for control 
fish while the lower half contains the data for experimental fish. The figures 
in the uppermost part of the table indicate the various high temperatures to which 
the fish were exposed. The equations under each temperature show the number 
of fish per sample which survived at that particular temperature for 24 hours. 
For example, in test #1 of the controls, 5 out of 6 fish survived at 30.5 C., 
6 out of 6 survived at 31.0 C., and none out of three survived at 31.5 C. when 



158 J. CHARLES CHEVERIE AND W. GARDNER LYNN 

left at these temperatures for 24 hours. In test #2 of the controls, three out of 6 
fish survived 24 hours' exposure at 31.0 C., while all of the 6 fish placed in 
31.5 C. died within 24 hours. Since the early tests indicated that the upper 
lethal temperature for the whitecloud was 31.0 C., the majority of the tests 
were made at this temperature. Totals of these tests show that 8 out of 12 fish, 
or 67%, survived 24 hours at 30.5 C., 26 out of 47, or 55%, survived 31.0 C, 
while 24-hour exposure to 31.5 C. proved lethal for all fish kept at this tempera- 
ture. Similar comparisons may be made for the experimental animals by examin- 
ing the lower part of the table. A comparison of the results obtained for controls 
and experimentals indicates that the controls had a greater capability to withstand 
high temperature than did the fish treated with thiourea. It should be mentioned 
that there was no significant difference in the survival in males and females. 

DISCUSSION 

Piscine thermal tolerance extremes depend on previous acclimation tempera- 
tures. This fact is evident from the numerous thermal studies done on fish since 
the time of Loeb and Wasteneys (1912). The rate at which acclimation occurs, 
however, seems to differ widely in different fish. Some authors (Doudoroff, 
1942; Brett, 1944, 1946) have found that gain of heat tolerance is more rapid 
than loss in the process of acclimation. Tsukuda (1960), on the other hand, 
reported that changes in cold tolerance are almost parallel to changes in heat 
tolerance in the guppy, Lebistes reticulatus, with both processes being relatively 
slow. Using approximately the same temperature change intervals Doudoroff 
(1942) and Brett (1944) found for Girella nigricans and Phimephales powelas, 
respectively, that a period of 20 days or more was required to completely acclimate 
these fish at low temperatures. About 35 days are required to acclimate male 
guppies at low temperatures (Tsukuda, 1960). Although no rate of acclimation 
was ascertained for the whitecloud in the present experiments, in view of the 
findings of Doudoroff, Brett and Tsukuda, and of the relative stability obtained 
in the lethal temperature for the whitecloud, the 40-60 days' exposure to 15.0 C. 
seems to have been a sufficient period to bring about a stable physiological condition 
at the low temperature. 

Under the conditions of this experiment the high-temperature death point for 
controls of the cyprinid, Tanichthys albonubes, was found to be 31.0 C. Brett 
(1956) tabulated the reported lethal temperatures for a number of species of fish. 
Members of the Cyprinidae recorded by Brett have upper lethal temperatures 
ranging from 28.9 C. to 32.8 C. Hence, the upper lethal temperature ascer- 
tained for the whitecloud here is in agreement with the range of upper thermal 
limits recorded for other members of the family. 

With a few exceptions, the thyroid of teleosts consists of unencapsulated fol- 
licles scattered individually or in small groups in the connective tissue along the 
ventral aorta and branchial arteries in the lower jaw of the fish. The thyroid 
morphology of Tanichthys albonubes is consistent with this general pattern. This 
diffuse nature of the teleost thyroid renders complete surgical extirpation of the 
gland impossible. To study the effect of hypothyroidism in these animals one must 
have recourse to chemical inhibitors or to thyroidectomy by radioiodine. Thiourea 
has been used on teleosts by a number of investigators (Scott, 1953; Fortune, 



THYROID AND TEMPERATURE TOLERANCE 159 

1953, 1955, 1956, 1958; Frieders. 1949, 1954; Suhrmann, 1955; Auerbach, 1957; 
Theobald, 1959; and others). Common effects of thiourea treatment on the thy- 
roid histology of fish include increased vascularization in the thyroid area, hyper- 
plasia and hypertrophy of the follicular epithelium, and a loss of stored colloid. 
The degree to which these results are achieved depends on the species of fish used, 
the concentration of thiourea, the length of the treatment and other factors. This 
hyperactive appearance of the thyroid after thiourea treatment may be explained 
on the basis of the thyroid-pituitary relationship which exists in fish as in other 
vertebrates. Thiourea prevents the thyroid from synthesizing its hormone so that, 
as soon as the supply of circulating hormone present at the beginning of thiourea 
treatment falls below a certain level, the pituitary begins to secrete thyrotrophin 
to activate the thyroid to produce more hormone. The thyroid, however, because 
of continued treatment, remains hypofunctional despite the fact that it is hyper- 
active. Nevertheless, in some species under certain conditions, prolonged treat- 
ment with antithyroid drugs may result in an "escape" from thyroid inhibition 
(Pickford and Atz, 1957). Frieders (1949) observed that the hyperplastic thy- 
roid of Trichogaster trichopterns resulting from 0.0025^ thiourea treatment at 
room temperature returned to a normal state after the fifth week of treatment. 
Similar reactions to thiourea were shown by Fortune (1958) to exist in Pho.vinus 
laevis. The hyperemic, hyperplastic and hypertrophic condition of the thyroid 
following thiourea administration in the present work is clear evidence that the 
thyroid of the whitecloud was quite responsive to thiourea treatment. Since this 
hyperactive state persisted from the beginning of the acclimation at low tempera- 
ture to the end of the experiment, it is concluded that there was no "escape" from 
the thyroid inhibition, and that the thyroid hormone was totally or to a great degree 
suppressed in the experimental animals during this time, a desired condition for 
the experiment. 

The dramatic results obtained by Fortune (1953, 1955) served as a basis for 
the hypothesis that decreased thyroid function in teleosts at high temperature is 
a significant factor in causing death at the lethal temperature. Fortune (1955) 
kept Phoxinus (pho.vinus) laevis in thiourea at 23 C. for three days and then 
subjected them to a rise in temperature to 33.0 C. over a period of two days. The 
experimental fish survived indefinitely at this increased temperature and appeared 
normal, while non-thiourea-treated fish died under such conditions between 23.0 C. 
and 24.0 C. Although no data were given, Fortune reported in the same paper 
that the thermal range of Lebistes rcticidatns could also be extended by thiourea 
treatment. 

Other experiments involving high-temperature tolerances and thyroid hypo- 
function in teleosts have produced conflicting results. Evropeitzeva (1949) reared 
six-day-old fry of Coregonus lavaratus hidoga for 17 days in thiourea at room 
temperature and then exposed 100 control and experimental animals to 29.0 C. 
for 5 minutes. Ninety-seven per cent of the control animals died while 97% of 
the experimental animals survived the 5-minute exposure. These results seem 
to agree with Fortune's hypothesis. The studies of LaRoche and Leblond (1954) 
and Auerbach (1957) on other young fish, however, oppose the work of Evro- 
peitzeva. LaRoche and Leblond thyroidectomized young Atlantic salmon, Saliuo 
salar, by radioiodine and found that thyroidectomy impaired the ability of these 



160 J. CHARLES CHEVERIE AND W. GARDNER LYNN 

salmonid parr to withstand rising temperatures (5.0 C. to 10.0 C.)> while the 
thyroidectomized fish receiving thyroid material in their diet survived such rising 
temperatures. Auerbach acclimated young Xiphophorus Jielleri at 15.0 C. and 
25.0 C., and then put some of the fish in 0.15% thiourea at the same temperatures 
for 14 days. She tested the control and experimental fish at high temperatures 
and found that the cold- and warm-adapted controls evidenced heat coma at higher 
temperatures than did the cold- and warm-adapted thiourea-treated fish. 

Using adult fresh-water fish, Suhrmann (1955) decreased thyroid function in 
Carassius vulgaris by keeping these fish in thiourea at 5.0 C. and 26.0 C., and 
then found that heat coma occurred in the cold- and warm-acclimated fish at 
31.4 C. and 35.2 C., respectively. By comparing her results with the findings 
of Christophersen and Precht (1952) for untreated Carassius vulgaris kept at the 
same acclimation temperatures, she deduced that thiourea increased upper tempera- 
ture tolerance for cold-acclimated fish, but decreased it for the warm-acclimated 
ones. 

The work of Auerbach (1957) and Theobald (1959) on other adult fresh-water 
fish does not agree with Fortune's results. Auerbach reported that controls for 
Lebistes reticulatus and Platy variatus acclimated at 25.0 C. and 15.0 C. had 
higher heat coma temperatures than experimental animals acclimated at the same 
temperatures but subjected to a 14-day treatment of 0.15% thiourea. Theobald 
administered various concentrations of thiourea or thyroid-stimulating hormone 
(TSH) to Gambusia affinis for 5 weeks at 25.0 C., and then acclimated the fish 
to 30.0 for one week. He found that the upper lethal temperatures for control, 
thiourea-treated, and TSH-treated animals were 37.4 C., 35.6 C. and 37.9 C., 
respectively. 

Auerbach (1957) reported that the marine fish Leuciscus rutilus, if treated 
with thiourea, could not stand as high temperatures as non-treated fish. She kept 
a group of this species at 5.0 C. for 2-4 weeks and a group at 20.0 C. for 1-2 
weeks and then put half of each group in 0.15% thiourea for 14 days at these 
temperatures. She found that cold-adapted controls had a heat coma temperature 
range of 28.1 C. to 29.0 C., while the cold-adapted thiourea-treated fish had a 
heat coma temperature range of 26.3 C. to 27.0 C. The warm-adapted controls 
and thiourea-treated fish had heat coma ranges of 29.7 C. to 30.8 C. and 27.6 C. 
to 28.4 C., respectively. 

The upper lethal temperature for whiteclouds exposed to 0.05% thiourea under 
the conditions of the present experiment was found to be 30.0 C., whereas the 
upper lethal temperature for untreated control fish was 31.0 C. Thus, these 
results are contrary to those of Fortune (1955) for Phox'mus and Lebistes, and 
indicate that thiourea treatment slightly decreased the ability of Tanichthys albo- 
nubes to withstand high lethal temperatures. 

SUMMARY 

1. The thyroid of Tanichthys albonubes was rendered hypofunctional by 30-40 
days' immersion in 0.05% thiourea. This treatment affected the thyroid histology 
to a significant degree, resulting in hyperemia, follicular hyperplasia, and cellular 
hypertrophy with a loss of stored colloid. 



THYROID AND TEMPERATURE TOLERANCE 161 

2. After each group had been acclimated at 15.0 C., the upper lethal tempera- 
tures were determined for the thiourea-treated and control fish. The upper lethal 
temperature of the controls was 31.0 C. and that of the experimentals was 30.0 C. 
This indicates that in this species decreased thyroid function does not result in 
increased high-temperature tolerance but, in fact, slightly decreases it. 

LITERATURE CITED 

AUERBACH, M., 1957. Hat die Schildriise fiir die Temperaturadaptation der Fische eine Be- 
deutung? Zeitschr. Fiscli. Hilfswiss., 6: 605-620. 

BLACK, E. C., 1953. Upper lethal temperatures of some British Columbia freshwater fishes. 
/. Fish. Res. Bd., Canada, 10 : 196-210. 

BRETT, J. R., 1941. Tempering versus acclimation in the planting of speckled trout. Trans. 
Amcr. Fish. Soc., 70 : 397-403. 

BRETT, J. R., 1944. Some lethal temperature relations of Algonquin Park fishes. Pub. Ont. 
Fish. Res. Lab., 63 : 1-49. 

BRETT, J. R., 1946. Rate of gain of heat-tolerance in goldfish (Carassius auratus). Pub. Out. 
Fish. Res. Lab., 64: 9-28. 

BRETT, J. R., 1952. Temperature tolerance in young Pacific salmon, genus Oncorhynchus. 
'/. Fish. Res. Bd., Canada, 9: 265-323. 

BRETT, J. R., 1956. Principles of thermal requirements of fish. Quart. Rev. Biol., 31 : 75-87. 

CHRISTOPHERSEN, J., AND H. PRECHT, 1952. Untersuchungen zum Problem der Hitzeresitenz. 
I. Yersuche an Karauschen (Carasshts vulgaris Nils.). Biol. Zentralbl., 71: 313-326. 

DOUDOROFF, P., 1942. The resistance and acclimatization of marine fishes to temperature 
changes. I. Experiments with Girella nigricans (Ayres). Biol. Bull., 83: 219-244. 

EVROPEITZEVA, N. V., 1949. Influence of thiourea on development of the thyroid gland in 
sigludoga (Coregonus lararetus ludoga Pol). Dokl. Akad. Nauk., USSR., 68: 977- 
980. (In Russian.) 

FISHER, K. C., 1958. An approach to the organ and cellular physiology of adaptation to tem- 
perature in fish and small mammals. In: Prosser, C. L. (ed.) ; Physiological Adapta- 
tion. New York, Ronald Press. 

FORTUNE, P. Y., 1953. Thyroid activity in teleosts. Nature, 171 : 483-484. 

FORTUNE, P. Y., 1955. Comparative studies of the thyroid function in teleosts of tropical and 
temperate habitats. /. Exp. Biol., 32: 504-513. 

FORTUNE, P. Y., 1956. An inactive thyroid gland in Carassius auratus. Nature, 178: 98. 

FORTUNE, P. Y., 1958. The effect of temperature changes on the thyroid-pituitary relationship 
in teleosts. /. Exp. Biol., 35: 824-831. 

FRIEDERS, F., 1949. The effect of thiourea and phenylthiourea on growth and pigmentation of 
several species of fish. Master's dissertation, The Catholic University of America. 

FRIEDERS, F., 1954. The effects of thyroid-inhibiting drugs on some tropical fish. Catholic 
Univ. Amcr. Biol. Stud., 31: 1-37. 

FRY, F. E. J., J. R. BRETT AND G. H. CLAWSON, 1942. Lethal limits of temperature for young 
goldfish. Rev. Canadienne dc Biol., 1 : 50-56. 

FRY, F. E. J., J. S. HART AND K. F. WALKER, 1946. Lethal temperature relations for a sam- 
ple of young speckled trout, Salvelinus fontinalis. Pub. Ont. Fish. Res. Lab., 66: 1-35. 

HATHAWAY, E. S., 1927. Quantitative study of the changes produced by acclimatization on 
the tolerance of high temperatures by fishes and amphibians. Bull. U. S. Bur. Fish., 
43: 169-192. 

LAROCHE, G., AND C. P. LEBI.OND, 1954. Destruction of thyroid gland of Atlantic salmon 
(Salino salar L.) by means of radioiodine. Proc. Soc. Exp. Biol. Mcd., 87: 273-276. 

LOEB, J., AND H. WASTENEYS, 1912. On the adaptation of fish (Fundulits) to higher tempera- 
tures. /. Exp. Zool., 12 : 543-557. 

PICKFORD, G. E., AND J. W. ATZ, 1957. The Physiology of the Pituitary Gland of Fishes. 
New Y 7 ork Zoological Society, New York. 613 pp. 



162 J. CHARLES CHEVERIE AND W. GARDNER LYNN 

PRECHT, H., 1958. Concepts of the temperature adaptation of unchanging reaction systems of 

cold blooded vertebrates. In: Prosser, C. L. (ed.) ; Physiological Adaptation. New 

York, Ronald Press. 
SCOTT, J. L., 1953. The effects of thiourea treatment upon the thyroid, pituitary and gonads of 

the zebra fish, Brachydanio rerio. Zoologica, 38 : 53-62. 
SUHRMANN, R., 1955. Weitere Versuche iiber die Temperaturadaptation der Karauschen 

(Carassius vulgaris Nils). Biol. Zentrabl., 74: 432-448. 
SUMNER, F. B., AND P. DouDOROFF, 1938. Some experiments on temperature acclimatization 

and respiratory metabolism in fishes. Biol. Bull., 74 : 403-429. 
THEOBALD, P. V. K., 1959. The relation of thyroid function to upper lethal temperature in 

Gambusia affinis. Catholic Univ. Amer. Biol. Stud., 50: 1-37. 
TSUKUDA, H., 1960. Temperature adaptation in fishes. IV. Change in the heat and cold 

tolerances of the guppy in the process of temperature acclimatization. J. lust. Poly- 

techn. Osaka City Univ., 11 : 43-54. 



PELAGIC LARVAE OF NERINIDES AGILIS (VERRILL) 1 

DAVID DEAN AND PHYLLIS A. HATFIELD 

Department of Zoology and Entomology, University of Connecticut, Storrs, Conn., 
and the Marine Research Laboratory, Noank, Connecticut 

While the authors were engaged in a study of polychaetous annelid reproduction 
and development, a larva similar to that of Nerinides tridentata, described by 
Hannerz (1956), began to appear in the plankton of the Mystic River estuary 
in June and continued to occur through August of 1962. N. tridentata has not 
been reported from the New England coast, but N. agilis is found in sandy beaches 
along the Atlantic seaboard (Verrill, 1873; Hartman, 1945; Carpenter, 1956). 
A three-day search of local beaches yielded but one young adult. Since ripe adults 
were not available for laboratory fertilization studies, development of this spionid 
polychaete has been described from specimens obtained from the plankton and 
reared in the laboratory through metamorphosis and until identification of the 
species was definite. 

MATERIALS AND METHODS 

Larvae were collected in a no. 10 plankton net in qualitative tows from the 
Mystic River estuary, Noank, Connecticut. Larvae believed to belong to the same 
species, but representing various stages of development, were separated from 
plankton samples, described, photographed and drawn. For examination, larvae 
were placed on microscope slides with small pieces of Saran Wrap used as cover 
slips (Dean and Hatfield, 1963). This was effective in quieting the larvae without 
apparent injury. Photographs showing the outline of an undistorted larva and 
revealing a certain amount of detail were obtained using a Polaroid camera with 
10-second film mounted on a phase microscope. Additional detail was provided 
through further photographs, descriptions and quick drawings of characteristic 
parts. Composite outline drawings were made from the photographs and descrip- 
tions. Following examination some of the larvae were transferred to 5-cm. funnels 
(Wilson, 1952) used as rearing vessels, and raised to metamorphosis. Funnels 
contained filtered sea water and sediment of mixed particle sizes ranging from 
53 to 590 ju. Funnels were immersed in a bath of running sea water, and the 
larvae were fed liver powder (Howie, 1958) once a week. Sea water in the 
funnels was changed daily. 

The sea water used was filtered through a 47 ^ Millipore filter and stored in 
a stoppered flask immersed in the same water bath as the funnels containing the 
larvae. Sediment was obtained from the Mystic River estuary, washed three 
times with distilled water, air-dried, sieved with standard screens to obtain the 
particle sizes desired and stored in stoppered jars. 

1 This study was supported by research grant G22549 from the National Science Foundation. 
Contribution No. 26 of the Marine Research Laboratory, Noank, Conn. 

163 



164 



DAVID DEAN AND PHYLLIS A. HATFIELD 



OBSERVATIONS AND DISCUSSION 



Pelagic larvae of N. agilis should be easily distinguishable from all other poly- 
chaete larvae occurring in marine plankton of the New England region. The most 
apparent characteristic features of all planktonic stages are the acutely pointed 
prostomium and the typical opaque dark-brown to black gut. The prostomium 
has a protrusible finger-like process at its anterior extremity which is retracted 
by means of two longitudinal bands of muscles. The muscles arise on the antero- 
lateral aspects of the mouth and insert at the anterior tip of the process. The 
latter is usually retracted when the larvae are preserved. In the late pelagic stages, 
viz., 14- to 16-setiger larvae, the ability to retract this organ diminishes. 




FIGURE 1. Ventral view of the 3-setiger larval stage of Nerinides agilis. Ciliation of the 

vestibule has been omitted. 



Two lateral ear-like processes, called the umbrella by Hacker, as translated by 
Gravely (1909), are present from the 3- to 16-setiger stages and bear the proto- 
trochal cilia. Four dark brownish-red eyes occur in almost a straight line and are 
present in all stages. Palp buds appear first in the 5-setiger stage and arise dorso- 
laterally from the umbrella. The palps grow slowly until about the 15- to 16-setiger 
stage, at which time their growth seems more rapid, a canal develops within them 
and their mobility increases. 

Swimming setae occur in paired dorsal and ventral bundles on the first segment 
behind the umbrella, and are present up to the 16-setiger stage. At the 16- to 
17-setiger stage swimming setae are lost, leaving shorter adult setae in their place. 
Setation on other segments consists of capillary notosetae on all segments and capil- 
lary neurosetae on setigers 1 through 11. On the twelfth setiger five hooded bi- 
dentate crotchets appear and continue on the segments posteriorly. 

The mouth opens into the posterior part of the vestibule. The latter is an 



PELAGIC LARVAE OF NERINIDES AGILIS 



165 



0.2mm 




FIGURE 2. Dorsal view of the 5-setiger stage, showing palp buds posterior to the umbrella 
and the beginning of pharynx pigmentation. Sensory cilia have been omitted. 

elongate depression on the antero-ventral aspect of the head region (Fig. 1) which 
is bordered posteriorly by the anterior part of the first setiger. 

The pharynx appears as a large dark-brown to black pigmented area extending 
from the region of the peristomium posteriorly through setiger 2. The character- 




FIGURE. 3 Lateral view of the 10-setiger stage. The short cilia of the neurotroch can be 
seen on setigers 1 and 2. Gastrotrochs are present on setigers 3-9. Sensory cilia have been 
omitted. 



166 



DAVID DEAN AND PHYLLIS A. HATFIELD 



istic reticulate pattern of the pigment of the pharynx begins to develop at the 
5-setiger stage (Fig. 2). The density of pigment increases with larval development 
until at the 14-setiger stage it appears as dense as that of the stomach-intestine. 
Posteriorly from the pharynx an unpigmented oesophagus, about one segment 
in length, leads to the densely pigmented stomach-intestine. The latter region has 
lateral segmental diverticula (Fig. 4). The most posterior portion of the gut 




0.2mm 



FIGURE 4. Fourteen-setiger stage in dorsal view. Nototrochs appear on setigers 1-11. 
Hooded crotchets appear in the neuropodia of the last three setigers. 

of young larvae is unpigmented (Fig. 1), while the gut in the region of the last 
few setigers of older larvae (Figs. 3, 4, 6) varies from partially to well pigmented. 
According to Hannerz (1956) the comparable region of N. tridentata is straight 
and unpigmented. In N. agilis, however, the gut is enlarged in the pygidial region 
and is pigmented (Figs. 3 and 4). In the present study the gut posterior to the 
last diverticula was ciliated. 

Ciliation of N. agilis larvae is difficult to observe, due to the dense pigment of 
the gut. Cilia are of more than one type. Those of the prototroch, telotroch and 



PELAGIC LARVAE OF NERINIDES AGILIS 167 

the gastrotroch on setiger 3 are long and coarse. Cilia that appear on the prosto- 
mium are quite different, being variable in length, small in diameter and infre- 
quent in movement. Wilson (1928) refers to this type as "sensory cilia." All 
other cilia in the larvae are short. Prototrochal cilia are discontinuous both dorsally 
and ventrally. The telotroch, however, consists of a completely encircling row of 
cilia. Gastrotrochs first appear in the 7-setiger larva and occur only on setigers 
3 through 9, even in subsequent stages. There is one continuous band of cilia 
across the ventral surface of setiger 3. There are two patches per segment on 




FIGURE 5. Anterior end of 17-setiger stage in dorsal view. The larva is ready to meta- 
morphose. The larval setae of setiger 1 have been replaced by adult setae, and pharynx pig- 
mentation is all but lost. Nototrochs and sensory cilia have been omitted. 

setigers 4 through 9. The two patches are close together and near the ventral 
midline on setiger 4 ; on setigers 5 through 9 the patches become increasingly 
farther apart, until on setiger 9 they are lateral in position (Figs. 3 and 4). 

The neurotroch (Fig. 3) makes its appearance between the 3- to 5-setiger 
stage. It extends from the well ciliated vestibule to setiger 2 and there ends in 
an inconspicuous ciliated pit. 

Nototrochs were not observed in stages earlier than 14 setigers. At this stage 
cilia seemed to extend across the dorsal surface of segments 1 through 11 in a 
continuous band (Fig. 3). On 16-setiger stages examined for ciliation, noto- 
trochs were not seen on setiger 1 but on setigers 2 through 16. In addition, these 
later stages showed the nototrochs to occur in three patches instead of one continu- 
ous band. 

Several morphological changes take place in the larvae prior to metamorphosis. 



168 



DAVID DEAN AND PHYLLIS A. HATFIELD 



The reticulate pigment pattern of the pharynx that is so characteristic of this larva 
begins to break down in the 16-setiger stage, and by the 17-setiger stage (Fig. 5) 
only a few dark spots smaller than the eyes remain. Adult setae are formed and 
accompany the swimming setae in the 16-setiger stage. The telotroch disappears 
and septa between the gut and body walls appear. Parapodia become more pro- 
nounced and buds of the dorsal and ventral postsetal lamellae appear. The dark 
pigmentation of the stomach-intestine begins to disappear progressively from the 
anterior toward the posterior region. Also, as the time of metamorphosis nears, 
the larval swimming setae are less firmly attached. 




FIGURE 6. Posterior end in dorsal view of the same 17-setiger larva. Dense pigmentation 
of the posterior gut remains but the telotroch has been lost. Nototrochs have been omitted. 

Metamorphosis occurs at the 17-setiger stage. Larvae were collected from the 
plankton up to but not exceeding the 17-setiger stage. The most advanced 17- 
setiger larvae had lost their telotrochs and larval setae, while their prototrochs, 
although still bearing functional cilia, were reduced to small lobes dorsal to the 
palpal bases (Fig. 5). All larvae reared successfully in the laboratory to benthic 
stages metamorphosed at the 17-setiger stage. Young benthic stages were reared 
to 19-setiger stages, at which time branchiae were formed on the second setiger. 

Larvae reared in the laboratory were identical to larvae freshly obtained from 
plankton samples when specimens having the same number of segments were com- 
pared. In fact, the only variation noted during the study was a very slight differ- 
ence in density of gut pigment between individuals. This was true of both reared 
specimens and those from the plankton. 

The combination of the acutely pointed prostomium, the number and arrange- 
ment of eyes, the presence of branchiae beginning on setiger 2 and the bidentate 
hooded crotchets in the neuropodial segments show beyond any doubt that the 
larval development described is that of Nerinides agilis. 



PELAGIC LARVAE OF NERINIDES AGILIS 169 

SUMMARY 

Larvae of a spionid polychaete, occurring in the plankton of the Mystic River 
estuary from June through August, were isolated from plankton samples, photo- 
graphed and drawn. Some of the isolated larvae were reared in the laboratory 
through metamorphosis and until the species was positively identified as Nerinides 
agilis (Verrill). Developmental stages of reared specimens agreed well with 3- to 
17-setiger stages obtained from the plankton. Metamorphosis occurs at the 17- 
setiger stage. The 3-, 5-, 10-, 14- and 17-setiger stages are figured and pertinent 
larval characteristics described. The protrusible, acutely pointed prostomium, 
together with the deeply pigmented gut, are diagnostic features of all planktonic 
stages of this species. 

LITERATURE CITED 

CARPENTER, D. G., 1956. Distribution of polychaete annelids in the Alligator Harbor area, 

Franklin County, Florida. Florida State Univ. Studies, No. 22: 89-110. 
DEAN, D., AND P. A. HATFIELD, 1963. A method for holding small aquatic invertebrates for 

observation. Turtox News, 41 : 43. 

GRAVELY, F. H., 1909. Studies on polychaete larvae. Quart. J. Micr. Sci., 53 : 597-627. 
HANNERZ, L., 1956. Larval development of the polychaete families Spionidae Sars, Disomidae 

Mesnil, and Poecilochaetidae N. Fam. in the Gullmar Fjord (Sweden). Zool. Bidr. 

Uppsala, 31 : 1-204. 
HARTMAN, O., 1945. The marine annelids of North Carolina. Duke Univ. Mar. Sta., Bull. 

No. 2: 1-54. 
HOWIE, D. I. D., 1958. Dried organic substances as food for larval annelids. Nature, 181 : 

1486-1487. 

VERRILL, A. E., 1873. Report upon the invertebrate animals of Vineyard Sound and the adja- 
cent waters, with an account of the physical characters of the region. Rep. U. S. Fish. 

Comm. for 1871-72: 295-778. 
WILSON, D. P., 1928. The larvae of Polydora ciliata Johnston and Polydora hoplura Clapa- 

rede. /. Mar. Biol. Assoc., 15 : 567-603. 
WILSON, D. P., 1952. The influence of the nature of the substratum on the metamorphosis of 

the larvae of marine animals, especially the larvae of Ophelia bicornis Savigny. Ann. 

Inst. Ocean. Monaco, 27: 50-160. 



' 




THE RELATION OF WHOLE-BODY I 131 UPTAKE TO THYROID 

ACTIVITY IN THE DEVELOPING DOGFISH, 

SCYLIORHINUS CANICULA (L.) 

SISTER MARIE THERESE DIMOND 

Department of Biology, Trinity College, Washington 17, D. C. 

Inhibition of thyroid activity by various types of goitrogens is commonly 
ascertained by the presence of histological changes in the gland, such as hyperplasia, 
hypertrophy, and vacuolization of colloid, which are attributed to increased 
secretion of thyrotrophic hormone (TSH). Nevertheless, there are species among 
amphibians, e.g., Diemyctylus viridescens (Adams, 1946; Dent and Lynn, 1958; 
Lynn and Dent, 1961), and elasmobranchs, e.g., Scyliorhinus canicula (Tinacci, 
1947; Olivereau. 1952; Dent and Dodd, 1961) and Squalus suckleyi (Pritchard 
and Gorbman, 1960), which have apparently a very low secretion rate of TSH 
or a very slow response of the thyroid to the pituitary hormone, for histological 
effects from administration of antithyroid substances are minimal. Studies with 
a radioactive isotope of iodine, I 131 , indicate, however, that the common goitrogens 
are just as effective in inhibiting organic binding of iodine in these species as 
in those in which marked histological changes are produced (Olivereau, 1952; 
Dent and Lynn, 1958; Pritchard and Gorbman, 1960), and that TSH administra- 
tion stimulates iodine-binding and, in high concentration and after prolonged 
treatment, enlargement of thyroid follicles also (Dent and Dodd, 1961). 

Class experiments with tadpoles of Rana pipiens which had been immersed 
in 0.01%, 0.03%, or 0.05% thiourea solution for two weeks had revealed a 
higher whole-body I 131 uptake than occurred in the controls. This phenomenon 
was ascribed to the presumed hyperplastic condition of the thyroid gland in the 
experimentals, although no histological examination was made (Bileau, 1956). 
Questions arose as to whether one could correlate whole-body iodine uptake with 
thyroid activity in animals in which goitrogens bring about thyroid inhibition 
only, without producing goiter, and also whether one could distinguish between 
different types of thyroid inhibitors by such means. The goitrogens commonly 
employed interfere with two mechanisms, vis., the concentration of iodine (thio- 
cyanate. perchlorate) and its organic binding (thiourea and its derivatives) (Pitt- 
Rivers and Tata, 1959). 

Another aspect of thyroid activity which is of current interest is its relation 
to temperature in poikilotherms. Reports on teleost thyroid activity at both high 
and low temperatures are conflicting (Leloup and Fontaine, 1960; Olivereau, 
1960). Possibly the only report on elasmobranch response is that of Dent and 
Dodd (1961), who found that for hatchlings of Scyliorhinus canicula, the amount 
of iodine bound by the thyroid is greater at a higher temperature. 

The present study is concerned with the relation of whole-body I 131 uptake 
to organic binding of I i:u by the thyroid and to the histological condition of the 

170 



EMBRYO DOGFISH THYROID AND IODINE 171 

gland, upon treatment with thiourea or perchlorate at different temperatures. 
The form chosen was the embryo of the spotted dogfish, Scyliorhinus canicnla (L.). 
The thyroid in this species, which lacks a larval stage, in common with that of 
amniotes active at birth or hatching (Dimond, 1954; Mitskevitch, 1957), shows 
signs of activity comparatively early in development (Vivien and Rechenmann, 
1954). 

MATERIALS AND METHODS 

Twenty-five dogfish embryos, all in the latter half of development, were 
obtained in July and August, 1960, from an outdoor tank at The Laboratory, 
Plymouth, England. The mature egg cases had been removed surgically from 
captured females, and the embryos were developing in running sea water. 

For this work the specimens were removed from the egg cases and kept, each 
one separately, in fingerbowls with approximately 100 ml. sea water (salinity 
33. 3% c ) which had been taken at a depth and was considered equivalent to filtered 
water. Thiourea or potassium perchlorate was added to the sea water of the 
experimentals to make a concentration of 0.05%. Series I was kept at room 
temperature, which varied from 18 to 23.5 C. In Series II and III, however, 
both controls and experimentals were divided into two temperature groups, one 
at room temperature and the other at 8 C. Of the six initial specimens of 
Series I, only two were alive after 11 days in the same medium, probably because 
of heavy bacterial growth in the water. A third living specimen (1C), taken 
from the aquarium, was removed from its egg case and added to the series just 
at the time of iodine administration. For Series II (10 specimens) and III 
(8 specimens), the medium, including the plain sea water of the controls, was 
renewed once a week for the fish kept in the cold and twice a week for those at 
room temperature. Only one fish in each of these last series died, both of them 
from the cold group. 

Several of the larger fish were measured at the beginning of the experiment. 
It was difficult to keep them extended, however. Since I did not wish to run the 
risk of injuring them by removing them from the water, most of the fish were 
left unmeasured. Final measurements were made on the preserved specimens. 

After 11, 13, or 21 days of goitrogen treatment, the specimens were exposed 
to a solution of Nal 131 in carrier-free Na 2 S 2 O 3 diluted in plain sea water or 
drug solution to give an activity of 3 juc./ml. for Series I, and 1.8-1.85 ^c./ml. 
for Series II and III. After 39 to 50 hours the fish were washed and placed in 
fresh goitrogen solutions or sea water. Then 24 hours later, after another 
washing, each fish was put into a 75- or 100-ml. beaker of sea water in a gamma 
Geiger well counter and a count of whole-body radiation made, 10 minutes for 
Series I, and three one-minute periods for Series II and III. The gamma Geiger 
detector consisted of a glass well about one-fourth inch thick, surrounded by five 
Geiger-Muller tubes, with the whole apparatus encased in heavy lead shielding. 
Its resolution time limited accuracy of counting to about 50,000 counts per minute. 
Background averaged 243 counts per minute. 

In Series I, after the whole-body counts had been made, the yolk sac was 
tied off and severed from the body and counts made of the fish without yolk and 
of the yolk alone. 



172 



SISTER MARIE THERESE DIMOND 



Immediately after the counting, the embryos of Series I and II were killed 
by immersion in Bouin's solution. Those of Series III were replaced in the 
same solutions they had been in since the cessation of the I 131 treatment, and were 
counted again the next day. The specimens at 8 C. were then killed, whereas 
those at room temperature were held for three days longer at which time they 
were counted and preserved. The thyroid glands were then dissected out, em- 
bedded in polyester or ester wax, sectioned at 46 p., and mounted serially on 
slides with 0.1% amylopectin solution (Steedman, 1957). Alternate slides were 
stained, the greater number of them with Gomori's chrome alum hematoxylin 
phloxin, and mounted, and the others were covered by Kodak Autoradiographic 
Stripping Film AR.10. After one to eight days the film-covered slides were 
developed and allowed to dry at room temperature. Later they were either stained 
with gallocyanin and metanil yellow (Bowie and Edmonson, 1960), and dried, 
then mounted with Xam and covered, or were mounted and covered without staining. 

TABLE I 

Series I. Whole-body uptake and thyroid binding of I 131 in developing Scyliorhinus canicula, 
after immersion in 0.05% thiourea solution for 11 days prior to exposure to 3 p 
I 131 for 43 hours and washing for 24 hours at room temperature (18-23.5 C.) 









Counts per minute, corrected for background 


Protocol 


Total length 


Thyroid 




number 


in mm. 


autoradiography* 














Entire fish 


Body alone 


Yolk sac alone 


C4 


86 


+ + 


53,663 


53,391 


4,798 


1C 


53 


+ + 


10,267 


6,818 


4,135 


T4 


80 





5,633 


5,671 


643 



* Two plus signs indicate a dense autoradiogram with stripping film ; one plus sign, a faint 
autoradiogram ; and a minus sign, no autoradiogram. 



RESULTS 

Tables I, II, and III summarize the greater part of the pertinent information 
on the effects of the antithyroid drugs and temperature. The whole-body I 131 
count cannot bear statistical analysis because of the different stages of develop- 
ment and probably also of thyroid activity, the small number of specimens in any 
one group, and the varying lengths of time between treatment and fixation. Also, 
since the autoradiographic film was exposed for different periods of time, no 
quantitative comparison of the amount of blackening over the thyroid gland is 
possible. Nevertheless, some patterns emerge which seem to deserve comment. 

First of all, though, explanations of certain individual points are needed for 
clarification. The fish had a large yolk sac connected to the body by a cord 
(Figs. 1, 2) except for specimen 3P (Figs. 3, 4), which would probably have 
made its way out of the egg case about the time it was exposed to I 131 (Ford, 
1921). Five specimens (1C, 13P, 18C, 19P, and 20T) had still functional 
external gills at the termination of the study (Fig. 1), whereas for eight others (9C, 
10P, 11T, 12C, 14T, 15C, 16P, and 17T) the gills atrophied during the experiment 
before the time of I 131 administration (Fig. 2). The other individuals were 



EMBRYO DOGFISH THYROID AND IODINE 



173 



TABLE II 

Series II. Whole-body uptake and thyroid binding of 7 131 in developing Scyliorhinus canicula, 
after immersion in 0.05% thiourea solution for 13 days prior to exposure to 1.8 /j.c./ml. 
1 for 39 hours and washing jor 24 hours at either room temperature 
(18-23.5 C.) or 8 C. 



Protocol 
number 


Total length in mm. 


Temperature 


Goitrogen 
treatment 


Thyroid 
auto- 
radiog- 
raphy* 


Counts per minute, 
corrected for 
background 


Initial 


Final 


7C 


70 


85 


Rm. 


none 


+ + 


40,504 


5P** 


70 


87 


Rm. 


perchlorate 




5,272 


6T** 


70 


88 


Rm. 


thiourea 




11,358 


9C 


50 


63 


8 C. 


none 


+ 


1,390 


8P 


65-70 


70 


8 C. 


perchlorate 





1,715 


10P*** 


50 


63 


8C. 


perchlorate 






19Pf 




43 


8 P 
o \^. 


perchlorate 





514 


4T 


75 


75 


00 p 

o \~. 


thiourea 





1,479 


14T 




54 


8 C. 


thiourea 





846 


20T 




49 


8C. 


thiourea 





603 



* See Table I. 

** Gland lost in processing. 
*** Died between 10 and 13 days, 
t Exposed to only 0.9 juc./ml. I 131 . 

further developed at the beginning of the study. The thyroid gland, nevertheless,, 
in even the smallest and presumably the least advanced individuals, had very 
definite follicular structure with colloid more or less vacuolated, and cuboidal 
cells varying in height even in a given gland. There are no striking histological 
differences between the controls and experimentals at 8 C. or between those of 
Series I and II (14 and 16 days of exposure to the goitrogens, respectively) at 

TABLE III 

Series III. Whole-body uptake and thyroid binding of I 131 in developing Scyliorhinus canicula, 

after immersion in 0.05% thiourea solution for 21 days prior to exposure to 

1.85 nc./ml. I 131 for 50 hours and washing for 24 hours 





Total length in mm. 






Thyroid 


Counts per minute, corrected 
for background and decay 


Protocol 




Tempera- 


Goitrogen 


auto- 




number 






ture 


treatment 


radiog- 










Initial 


Final 






raphy* 


Initial 


21.5 hrs. 
later 


93 hrs. 
later 


15C 




80 


Rm. 


none 


+ + 


20,374 


16,929 


4,642 


3P 


85 


98 


Rm. 


perchlorate 





8,016 


5,841 


2,833 


16P 




85 


Rm. 


perchlorate 





1,987 


1,633 


1,223 


17T 




76 


Rm. 


thiourea 





5,488 


3,725 


1,375 


12C** 




66 


8 C. 


none 










18C 




54 


8 C. 


none 


+ 


1,084 


761 




13P 




60 


8 C. 


perchlorate 





1,282 


1,016 




11T 




66 


8 C. 


thiourea 





1,272 


930 





* See Table I. 
** Moribund and preserved at 21 days. 



174 



SISTER MARIE THERESE DIMOND 



room temperature. Nevertheless, somewhat higher cells, larger follicles, and more 
mitotic figures appear in the experimentals of Series III kept at room temperature 
(Figs. 5, 6, 7). 

Growth of the developing fish was generally slowed dow T n in the cold (Tables 
II and III), but not affected to any appreciable extent by either of the goitrogens. 
Exceptions to this, however, were specimens 9C and 10P, both of which increased 
in total length from 50 to 63 mm. during the cold-treatment. Activity was also 




FIGURE 1. Specimen 18C, 54 mm., kept at 8 C. for 21 days, at the time of exposure to 
I 131 . Notice the large yolk sac and the functional external gills (G). 

FIGURE 2. Specimen 15C, 80 mm., kept at room temperature. Notice a small yolk sac, 
the absence of external gills and the beginning of pigmentation. 

FIGURE 3. Specimen 3P, 98 mm., kept in 0.05% KQCh at room temperature. It is 
probably a hatchling. 

FIGURE 4. Ventral view of anterior region of specimen 3P. Notice the very small 
remnant of the yolk sac (Y). 

diminished at a low temperature, the fish remaining very quiet as compared to 
those at room temperature, but there was no evidence of goitrogen influence 
on behavior. 

In view of the size variations in the specimens, the I 131 whole-body counts are 
probably not significantly different for the control and experimental animals kept 
at 8 C. In the case of those held at room temperature, however, an obvious 
pattern is indicated, for in the three sets (omitting specimen 3P, which was at 
least 10 mm. longer than the others and probably a hatchling at the time), the 



EMBRYO DOGFISH THYROID AND IODINE 175 

relation of I 131 uptake was control > thiourea > perchlorate. The thyroid gland 
in the goitrogen-treated fishes, although it may have concentrated iodine, did not 
bind it, for there was no blackening of the autoradiographic film. In the controls at 
room temperature, however, the colloid of the gland appears well stocked with 
organic iodine (Figs. 8, 9), and even in those at 8 C., there is a sparse granulation 
of the autoradiographic film, indicating a low degree of iodine-binding (Figs. 
10, 11). 



'' v " " 6 




FIGURE 5. Section of thyroid gland of specimen 15C, 80 mm., kept at room temperature 
for 28 days. Cut at 6 JJL. Stained with chrome alum hematoxylin phloxin. Compare with 
Figure 2. 

FIGURE 6. Section of thyroid gland of specimen 16P, 85 mm., kept in 0.05% KC1O 4 at 
room temperature for 28 days. Staining and sectioning as in Figure 5. Notice the enlarged 
follicles, higher cells, and mitotic figures. 

FIGURE 7. Section of thyroid gland of specimen 17T, 76 mm., kept in 0.05% thiourea at 
room temperature for 28 days. Notice the same characteristics as in Figure 6. 

Finally, there is a marked difference in I i:n excretion between the control 
and goitrogen-treated fish, as seen in Figure 12. Iodine removal from the control 
occurred at an arithmetic rate, whereas for the fishes exposed to perchlorate and 
thiourea the curve approached an exponential pattern. 

DISCUSSION 

Iodine-131 in a concentration of 1.0 /xc./ml. for 24 hours has been found 
suitable for absorption by fresh-water forms (Lynn and Dent, 1957). Since sea 
water contains so much iodine (50 /xg./kg.), it was considered advisable to in- 
crease the amount of the radioactive form, and also the exposure time, in order to 
insure a measurable uptake. Actually, 1.0 /xc./ml. would probably have been 



176 



SISTER MARIE THERESE DIMOND 







,~\' ; ^\y':' 



:% 




.'*. 
* 



*' N < '*!* 

. -!.jto. "fl" < 



8 




4 * 

A 

r\ 

ikY* 



10 



it: 

FIGURE 8. Autoradiogram from longitudinally sectioned thyroid of specimen 15C, 80 
mm., kept at room temperature for 28 days. Compare with Figures 2 and 5. The animal was 
washed for 5 days after 50 hours' exposure to 1.85 yuc./ml. I 131 and then fixed. Four days 
later stripping film was applied to the sections and, after two days of exposure, developed. 
Cut at 6 IJL. 

FIGURE 9. Enlargement of a portion of Figure 8. 

FIGURE 10. Autoradiogram from longitudinally sectioned thyroid of specimen 18C, 54 
mm., kept at 8 C. for 25 days. The animal was washed for two days between I 131 treatment 
and fixation. Four days later stripping film was applied to the sections and developed the 
next day. Compare with Figure 1. Same magnification and sectioning as Figure 8. 

FIGURE 11. Enlargement of a portion of Figure 10. Same magnification as Figure 9. 



EMBRYO DOGFISH THYROID AND IODINE 



177 



satisfactory, for specimen 19P, not only the smallest of the fish, but also one 
treated with perchlorate, had a count almost three times background after exposure 
to only 0.9 /AC. /ml. The concentration of both thiourea and perchlorate (0.05%) 
was apparently optimal for Scyliorhinus canicula, for there was no evidence of 
toxicity effects, and the thyroid uptake of I 131 seemed to be completely inhibited. 
In Scyliorhinus canicula, thyroid activity begins very early in development. 
Vivien and Rechenmann (1954) report that at 25 mm., when the thyroid is still 



20,000 



I 5,000 



z 
I 



I 0,000 



12 

z 

D 
O 

o 



5,000 



ISC 




I 



2 
DAYS 



FIGURE 12. Elimination of I 131 from four embryos of Scyliorhinus canicula which had been 
kept at room temperature for 21 days prior to exposure to 1.85 /ic./ml. I 131 for SO hours and 
washing for 24 hours (Series III), as seen one and four days after the initial whole-body 
count. Counts per minute are corrected for background and decay. 

in the cord stage, it already binds iodine, but its role in embryonic life of elasmo- 
branchs remains undetermined. Among the multiple effects of thyroid hormone 
during development are both gross changes, such as yolk sac retraction in teleosts 
and sauropsids (Dales and Hoar, 1954; Dimond, 1954), and histological changes, 
such as degeneration of the pronephros, as observed in amphibians (Lynn and 
Peadon, 1955). 

In the present investigation, there were no obvious external differences between 
the controls and experimentals at a given temperature. Since only one specimen 
"hatched," no comparison can be made as to its relative speed of yolk-sac retraction, 



SISTER MARIE THERESE DIMOND 

and the histological examination was limited to the thyroid. It seems reasonable 
to presume, however, that some morphological and physiological differences could 
be found by detailed study, since the thyroid gland is obviously very active 
during normal development. 

Two facts concerning thyroid activity in the developing spotted dogfish at a 
low temperature (8 C.) stand out, viz., that the gland concentrates and binds 
iodine to a degree sufficient for autoradiographic determination, and that thyroid 
hormone formation is so low as to evade detection in a whole-body count (Tables 
II, III; Figs. 10, 11). Whether thyroid action here is limited to a low rate of 
hormone production and storage, or whether release and distribution to the rest 
of the body occur as well, is undetermined. Dent and Dodd (1961) report a 
four-fold increase in tail muscle iodine of hatchlings with a change of temperature 
from 8.6 C. to 13.6 C. 

Thiourea seems to have its typical action, viz., prevention of organic binding 
of iodine, with little or no effect on iodine concentration, for although there was 
no reaction on the stripping film, yet the whole-body count in the case of the 
fish at room temperature was markedly higher than that of the perchlorate-treated 
specimens. An attempt was made to determine the relative amounts of thyroidal 
iodide by freeze-drying the glands, but an accident to the vacuum pump prevented 
completion of the processing. 

The rate of thyroid hormone formation by dogfish embryos, and thus the 
histological response of the gland to thyroid inhibitors, appears to be directly 
related to the ambient temperature as well as to the state of development. 
Specimen 1C at room temperature, for example, had a whole-body count ten 
times greater than that of specimens 9C and 18C at 8. Of course, it had been 
exposed to a somewhat higher concentration of I 131 , but it is doubtful that that 
alone could account for the great difference. 

Further, although both thiourea and potassium perchlorate prevented iodine 
binding, yet in only Series III at room temperature, exposed to the drugs for a 
total of 28 days, were the typical histological responses of hypertrophy and hyper- 
plasia visible (Figs. 6, 7). Sixteen days' exposure (Series II) was not sufficient. 
Apparently even an active thyroid is slow as compared to that of homoiotherms. 

Pritchard and Gorbman (1960) found about a 25 C / increase in cell height 
and "vacuolization" of colloid, as well as low I 131 uptake in thyroid glands of 
near-term pups of Sqiialus snckleyi, after repeated injections with propylthiouracil. 
Tinacci (1947) observed some thyroid hyperplasia as well as hyperemia in the 
thyroid of Mustcliis lacvis after 25 days of oral administration of three different 
goitrogens, including thiouracil. On the other hand, Olivereau (1952), who 
administered thiourea or thiouracil to adults of Scyliorhinns canicula at 20 C., 
reports no histological differences between the controls and experimentals, other 
than the presence of almost pycnotic nuclei in the thyroid epithelium after 44 
injections of thiouracil. 

The pituitary hormone, however, is present and active in Scyliorhinus canicula, 
for hypophysectomy of developing pups prevents I 131 fixation (Vivien and Rechen- 
mann, 1954), and injection of mammalian TSH increases its organic binding in 
both hatchlings (Dent and Dodd, 1961) and adults (Leloup and Fontaine, 1960). 
Dent and Dodd have further shown that injection of ventral lobe extract from 



EMBRYO DOGFISH THYROID AND IODINE 179 

adult dogfish pitnitaries over a period of three weeks results in an enhanced I 131 
binding, although the histological picture is not changed. Nevertheless, after 
three weeks of treatment at 13-14 C. with a very large amount of mammalian 
TSH, at least five times that effective in man on a weight basis, the hatchling 
thyroid follicles are markedly enlarged. 

The results of investigations of temperature effects on reptilian thyroids 
substantiate the observations on elasmobranchs. Eggert (1936) observed that 
the thyroids of lizards kept at 6-7 C. during the summer are not activated by 
TSH, whereas at normal summer temperatures, the pituitary hormone is very 
effective. Shellabarger et al. (1956), working with turtles, found that TSH 
caused increased I 131 uptake at 21-23 C., but not at 2-3 C. The body tempera- 
ture of birds and mammals, whose thyroids are more responsive, is much above 
even the peak temperature in all these instances. 

Of course, other factors have to be considered once one leaves the realm 
of development, such as light intensity and periodism, osmotic conditions, seasonal 
cycles, degree of maturity, sexual activity (Eggert, 1936; Bileau, 1956; Shel- 
labarger et al., 1956; Hickman, 1959; Leloup and Fontaine, 1960; Olivereau, 
1960). The absence of thyroid hypertrophy and hyperplasia that Olivereau 
(1952) reports for adult Scyliorhinus canicula at 20 C. after prolonged treatment 
with thiouracil could be due to normal lack of utilization of thyroid hormone 
at the season of the year (autumn) or stage of life, or even under the prevailing 
conditions of light. Tinacci's (1947) work with Mnstclns laci'is was carried on 
during the summer, w r hich seems to be a time of high thyroid activity in poikilo- 
therms (Eggert, 1936; Bileau, 1956; Baggerman, 1957). 

It may be objected that Dent and Dodd (1961) observed thyroid response 
to TSH in the spotted dogfish during the winter. Their specimens, however, 
were hatchlings. Zezza (1937) points out that the thyroid gland of Torpedo 
ocellata presents a much more active histological picture during growth than 
in adulthood, and the same thing probably holds true for Scyliorhinus canicula. 
At any rate, the gland is very active during development. 

The rate of I 131 removal in Series III at room temperature shows an excellent 
correlation with both whole-body uptake and thyroid autoradiography. Specimen 
15C, with an arithmetic decrease in I 131 , must have had a constant turnover rate 
of thyroid iodine and consequent active excretion after exposure to an excess of 
iodide. On the other hand, in the three goitrogen-treated specimens (3P, 16P, 
and 17T), diffusion seems to have been the principal mechanism, for iodine 
removal occurred almost geometrically. Probably the deviation from exact ex- 
ponential change is due to the fact that the medium remained unchanged and I 131 
could diffuse back into the animals. Specimen 16P, with the lowest uptake, 
would be most affected by re-entrance of I 131 from the medium, and would 
reach equilibrium sooner. Unfortunately, water samples were not counted. 

The selachian embryo is a very satisfactory experimental animal, as Vivien 
(1954) has pointed out. Not only the oviparous forms, such as Scyliorhinus 
canicula, but also the ovoviviparous series, e.g., Squalus sncklcyi (Pritchard and 
Gorbman, 1960) can be reared in the laboratory. 

Many problems concerning thyroid activity in poikilotherms remain unsettled. 
Perhaps the study of the embryo, which only gradually develops its homeostatic 



180 SISTER MARIE THERESE DIMOND 

mechanisms, would reveal the basic function of the thyroid gland, which is then 
modified as development proceeds and environmental response becomes more 
complex. 

This investigation was supported in part by a PHS research grant, A-1766, 
from the National Institute of Arthritis and Metabolic Diseases, Public Health 
Service. The experimental portion of the work was carried out during the tenure 
of a National Science Foundation Faculty Fellowship, 69021, at the Laboratory 
of the Marine Biological Association of the United Kingdom, Plymouth, England. 
I am especially indebted to Dr. F. S. Russell, Director of The Laboratory, and 
to Dr. T. I. Shaw and Dr. D. B. Carlisle, under whose direction the experiments 
were performed. 

SUMMARY 

1. Embryos of Scyliorhimis canicula, removed from their cases and kept at 
either room temperature or 8 C, were treated with 0.05% thiourea or perchlorate, 
or they were untreated, for 11, 13, or 21 days previous to addition of I 131 to their 
medium. After two days of exposure to I 131 and a day of washing, whole-body 
counts were made. The thyroid glands were removed within four days and 
prepared for autoradiography. 

2. At 8 C., although the film indicated the presence of a slight amount of 
bound iodine in the controls, there was no marked difference in whole-body count 
in the three groups. 

3. At room temperature, the controls had very high counts and well-blackened 
autoradiograms, whereas the two experimental groups had low counts and no 
sign of organic iodine. 

4. Only the experimentals which were exposed to the goitrogens for 28 days 
showed histological responses, such as enlarged follicles and very numerous 
mitotic figures. 

5. Growth was inhibited at the low temperature, but apparently unaffected by 
either of the goitrogens at either temperature. 

LITERATURE CITED 

ADAMS, A. ELIZABETH, 1946. The effects of thiourea on the thyroids of Tritunis viridcscens. 
Anat. Rec., 94: 532. 

BAGGERMAN, BERTHA, 1957. An experimental study of the timing of breeding and migration 
in the three-spined stickleback (Gasterosteus aculeatus L.). Arch. Neerl. Zool., 12: 
105-318. 

BILEAU, SISTER M. CLAIRE, 1956. The uptake of I 131 by the thyroid gland of turtles after 
treatment with thiourea. Blol. Bull., Ill : 190-203. 

BOWIE, D. J., AND V. R. EDMONSON, 1960. Gallocyanin as a nuclear stain in autoradiography. 
Stain Technol., 35 : 1-4. 

DALES, S., AND W. S. HOAR, 1954. Effects of thyroxine and thiourea on the early develop- 
ment of chum salmon. Canad. J. Zool., 32 : 244251. 

DENT, J. N., AND J. M. DODD, 1961. Some effects of mamalian thyroid stimulating hormone, 
elasmobranch pituitary gland extracts and temperature on thyroidal activity in newly 
hatched dogfish (Scyliorhlnus caniculus}. J. EndocrinoL, 22: 395-402. 

DENT, J. N., AND W. G. LYNN, 1958. A comparison of the effects of goitrogens on thyroid 
activity in Triturus viridescens and Desmognathus fuscns. Biol. Bull., 115: 411-420. 



EMBRYO DOGFISH THYROID AND IODINE 181 

DIMOND, SISTER MARIE THERESE, 1954. The reactions of developing snapping turtles, 
Chelydra scrpentina serpcntina (Linne), to thiourea. /. Exp. Zoo/., 127: 93-116. 

EGGERT, B., 1936. Zur Morphologic und Physiologie der Eidechsen-Schilddriise. II. "Qber 
die Wirkung von hohen und niedrigen Temperaturen, von Thyroxin und von thyreo- 
tropem Hormon auf die Schilddriise. Zeitschr. iwss. Zoo/., 147 : 537-594. 

FORD, E., 1921. A contribution to our knowledge of the life-histories of the dogfishes landed 
at Plymouth. /. Mar. Blol. Assoc., 12: 468-505. 

HICKMAN, C. P., JR., 1959. The osmoregulatory role of the thyroid gland in the starry 
flounder, Platichthys stellatus. Canad. J. Zoo/., 37: 997-1060. 

LELOUP, J., AND P. FONTAINE, 1960. Iodine metabolism in lower vertebrates. Ann. N. Y. 
Acad. Sci., 86: 316-353. 

LYNN, W. G., AND J. N. DENT, 1957. Phenylthiourea treatment and binding of radioactive 
iodine in the tadpole. Biol. Bull., 113: 160-169. 

LYNN, W. G., AND J. N. DENT, 1961. A comparison of the responses of Tritiirus and 
Desmognathus to thyroid-stimulating hormone administration. Biol. Bull., 120: 54-61. 

LYNN, W. G., AND ANNIE M. PEADON, 1955. The role of the thyroid gland in direct develop- 
ment in the anuran, Eleuthcrodactylus martinicensis. Growth, 19: 263-286. 

MITSKEVICH, M. S., 1957. Glands of internal secretion in the embryonic development of 
birds and mammals. Published for the National Science Foundation and the Depart- 
ment of Health, Education and Welfare, U. S. A., by the Israel Program for 
Scientific Translations, 1959. 304 pp. 

OLIVEREAU, MADELEINE, 1952. Action de divers antithyroidiens sur la destinee de 1'iode 
radioactif 131 I dans la glande thyroide d'un Selacien (Sc\llinm canicula L.). C. R. Soc. 
Biol. (Paris}, 146: 569-570. 

OLIVEREAU, MADELEINE, 1960. Quelques aspects anatomiques et physiologiques de la glande 
thyroide chez les poissons. Ann. Soc. Zoo/. Bclg., 90: 83-98. 

PITT-RIVERS, R., AND J. R. TATA, 1959. The Thyroid Hormones. The Pergamon Press, 
London. 247 pp. 

PRITCHARD, A. W., AND A. GoRBMAN, 1960. Thyroid hormone treatment and oxygen 
consumption in embryos of the spiny dogfish. Biol. Bull., 119: 109-119. 

SHELLABARGER, C. J., A. GORBMAN, F. C. SCHATZLEIN AND D. McGiLL, 1956. Some quantita- 
tive and qualitative aspects of iodine-131 metabolism in turtles. Endocrinol<><i\, 59: 
331-339. 

STEEDMAN, H. F., 1957. Polyester wax : a new ribboning embedding medium for histology. 
Nature, 179: 1345. 

TINACCI, F., 1947. L'azione de alcune sostanze antitiroidee (tiuracile, metiltiuracile e aminoti- 
azolo) sul Mustelus laevis. Pubbl. Staz. Zoo/. Napoli, 21 : 124-131. 

VIVIEN, J. H., 1954. Quelques exemples de 1'utilisation du germe et de 1'embryon de Selacien 
dans les recherches experimentales concernant la regulation, les parageneses, 1'organo- 
genese et la physiologic embryonnaire. Arch. Anat. Histol. et Embryol. Strasbourg, 
37: 163-174. 

VIVIEN, J., AND R. RECHENMANN, 1954. Etude sur la fonction thyroidienne de 1'embryon de 
Selacien. C. R. Soc. Biol. (Paris), 148: 170-172. 

ZEZZA, P., 1937. Tiroide, maturita sessuale e gestazione in Torpedo occllata. Boll. Soc. Ital. 
Biol. Sper., 12: 74-76. 



RESISTANCE OF THE PURPLE SEA URCHIN TO OSMOTIC STRESS 1 

A. C. GIESE AND A. FARMANFARMAIAN 2 

Hopkins Marine Station, Pacific Grove, California, and Department of Biological Sciences, 

Stanford University, Stanford, California 

The western purple sea urchin, Strongylocentrotus purpuratus, is briefly ex- 
posed to drastic environmental changes during the very low tides which occur in 
summer and winter along the central California coast. On such occasions these 
animals may be subjected to dehydration, due to high temperature, direct exposure 
to the sun, and particularly the wind. During heavy rains at low tides they may 
be exposed to considerably diluted sea water. The sea urchin is thus exposed to 
a wide range of salinities in the intertidal area. The osmotic tolerance of these 
animals was therefore determined by observations and experiments reported in 
this paper. During winter months when tide-pool water is diluted by rains, spawn- 
ing may be observed ; therefore, the influence of salinity changes on fertilization 
and development was also examined. The results reported in this paper suggest 
that sea urchins and their developing embryos can tolerate considerably greater 
variations in osmotic conditions than they are likely to meet in their environment, 
indicating an ample "safety factor" in their constitution. 

MATERIALS AND METHODS 

The sea urchins were collected at low tides, primarily near Yankee Point, five 
miles on the Pacific Coast south of Carmel, California. At this point large popu- 
lations of the purple sea urchin are exposed at low tides. Other collections were 
made at Moss Beach, California, near Stanford University. At the Hopkins Marine 
Station the urchins were put into aquaria with running sea water at a temperature 
varying between 12 and 16.7 C. At the University they were kept in a constant 
temperature room at 13 C. in well-aerated sea water, changed daily, until the 
specimens were used. 

The resistance of the adult sea urchins to changes in salinity had to be assessed 
by their reactions and appearance. Healthy sea urchins, if not too crowded, tend 
to crawl up on the sides of aquaria rather than stay at the bottom, while unhealthy 
animals usually remain on the bottom. Healthy urchins eat algae avidly, sickly 
ones do not. Healthy urchins, when overturned, right themselves in a coordinated 
manner within about a minute, whereas unhealthy ones may fail to do so, or take 
more time. Stimulation of normal animals with a probe or a bright spot of light 
(e.g., from an American Optical Company Universal Illuminator) brings about 
a positive reaction (local erection, in the direction of stimulation, of spines and 
pedicellariae ) and stronger stimulation elicits an escape reaction. These responses 
are abnormal in unhealthy animals. Finally, animals which are sick or dying 

1 Supported in part by U.S.P.H.S. Grant #RG-4578. 

2 Now at Pahlavi University, Shiraz, Iran. 

182 



OSMOTIC RESISTANCE OF A SEA URCHIN 183 

lose pigment and spines. Detection of slight changes in appearance or behavior 
of these animals is somewhat subjective, but marked changes are very definite and 
easy to note. Failure to show any response to prodding indicates that an animal 
is, to all intents and purposes, dead. 

It proved impossible to develop a consistently reliable source of running sea 
water at various salinities because of difficulties with regulating valves and air 
embolism in the tubes. The various solutions to be tested were made up in 7-liter 
battery jars, using appropriate mixtures of sea water, tap water and saturated 
brine (from the salt pools at Moss Landing, California), and aerated vigorously. 
Eight to 10 animals, 3.0-3.5 cm. in test diameter and totaling about 200 grams wet 
weight, were kept in each jar. The concentrations of sea water tested were 61.5%, 
70.4%, 80.7%, 90.3%, 110%, 121.4%, 130.8%, 150% and 170% as determined 
by titration with silver nitrate, using potassium chromate as indicator. 

In the developmental studies at different salinities the following procedures were 
observed : Ovaries were removed from gravid female sea urchins, avoiding con- 
tamination with other tissues, and suspended in filtered sea water in Syracuse 
watch glasses. The concentrated egg suspension was picked up in a mouth pipette 
under a dissecting microscope and added, with a minimum of sea water, to about 
1 ml. of the appropriate hypertonic or hypotonic sea water (made up with distilled 
water). Sperms were added in small quantities from the tip of a needle, and 
the progress of fertilization, first and second cleavages, early and late blastulation, 
gastrulation, and in some instances, pluteus formation, was noted. 

Changes in the osmotic pressure of the perivisceral fluid of the sea urchins 
exposed to sea water of various tonicities were determined by the method of Gross 
(1954). The effects of tonicity changes of sea water on the respiration of sea 
urchins were studied with the standard Warburg-Barcroft manometric method, 
using large respirometric flasks which readily accommodate small sea urchins. 

In all experiments small sea urchins were employed in order to have a more 
homogeneous population sample. For respirometry they had to be 2-3 cm. in 
diameter to fit into the flask (Farmanfarmaian, 1959). In the other experiments 
specimens 3.0-5.5 cm. in diameter were used. Experiments were carried out soon 
after the animals were collected and when they were all well fed and healthy. Feed- 
ing was avoided during experiments (except when testing the feeding reaction) 
to prevent fouling the aquaria. The latter had to be cleaned frequently at first 
because of extensive defecation. When an animal died it was promptly removed, 
since all the other animals in a container will soon die unless this is done. 

EXPERIMENTAL RESULTS 
1. Osmotic tolerance of sea -urchins 

It is apparent from the results in Table I that the extreme salinities are almost 
immediately injurious; no response to stimulation was obtained after a three-hour 
exposure of sea urchins to 30%, 50%, 150% and 170% sea water. Those in 60% 
and 130% sea water, which also showed no response after three hours, recovered 
when replaced in sea water. Most of the studies were concerned with the other 
concentrations, close to sea water, because they lie within the range of greater 
ecological interest. 



184 



A. C. GIESE AND A. FARMANFARMAIAN 



TABLE I 
Osmotic tolerance of adult Strongylocentrotus purpuratus* 



Per cent 
sea water 


Effect of short exposure 
(3 hrs.) 


Effect of prolonged exposure 
(days) 


30 


No activity or response ; no 
recovery (die) 





50 


No activity or response ; 
recover later* 1 


Die by second day. 


60 


Little activity or response; 
recover later*' 


Lose much pigment ; unresponsive ; most 
by the second day. 


of them die 


70 


Normal at beginning 


Stay near bottom of container; some die 
day, all by 25th day. 


on 5th 


80 


Normal*** 


Lose some pigment for first three days; survive")" 
35 days. 


90 


Normal*** 


Lose a little pigment for first three days 
35 days. 


survive")" 


100 


Normal*** 


Normal ; survive f 35 days. 


110 


Normal*** 


Lose some pigment continuously; some reduction 
in activity ; survive f 35 days. 


120 


Normal at beginning, reduced 
response; recover later** 


Lose pigment, stay near bottom of container; some 
dead by 7th day; half are dead by the 35th day. 


130 


Little activity or response; 
recover later** 


Lose pigment ; die on second day. 


150 


No activity or response ; 
no recovery 





170 


No activity or response ; 
no recovery 






* Temperature varied between 12.0 and 16.7 C. during the course of these experiments. 
** When replaced into sea water (100%). 
*** For the entire period of observation 35 days. 

t Climb up sides of container (at least at first) ; right themselves rapidly ; feed upon Iridaea 
(red alga) ; respond to bright light and touch. 

Although the sea urchins tolerated 70% and 120% sea water for three hours, 
they were not normal after a more prolonged exposure and they stayed near the 
bottom of the tank instead of climbing along the sides as did the controls. They 
then lost pigment and many died in 70% and 120% sea water between the twenty- 
fifth and thirty-fifth day of exposure. These concentrations of sea water, there- 
fore, constitute the limits of tolerance. Tonicities of 80%, 90% and 110% sea 
water were tolerated, the animals remaining essentially normal for at least 35 days 
(in some cases to 50 days) of exposure and observation. Needless to say, controls 
in sea water remained normal for much longer periods. 



OSMOTIC RESISTANCE OF A SEA URCHIN 



185 



The response to changes in the concentration of sea water is tolerance and not 
regulation. This is indicated by the change in weight of sea urchins following 
immersion in sea water of diverse tonicities, as seen in Figure 1A for one series 
of experiments. The animals used were 3.0 to 5.0 cm. in test diameter and weighed 
25 to 35 grams. They were immersed in 500 ml. of the appropriate solution in 
a 600-ml. beaker and drained on a towel for 5 minutes before weighing. The 
change in \veight was calculated as per cent of original weight. Figure IB shows 
the change in the osmotic pressure of the perivisceral fluid as a result of immersion 

A B 



101 r 



Returned to SW 




.7 



.6 



Returned to SW 



,5 



Returned to SW 



100 




Perivisceral fluid 



50 



75 



100 



125 



Hours Per cent sea water 

FIGURE 1. A. Changes in weight of sea urchins placed in various concentrations of sea 
water and again after replacement in sea water. B. Changes in concentration of perivisceral 
fluid at equilibrium (after a 1.5-hour exposure of the sea urchin to the solution) ; concentration 
is given in terms of NaCl equivalents. Note that the perivisceral fluid approximates the 
external bathing medium in concentration. 



186 



A. C. GIESE AND A. FARMANFARMAIAN 



c 
o 

"o. 

E 



C 

o 
o 

C 
CD 

cn 

X 

o 

CD 



CD 

cr 




100 



200 



300 



400 



Per cent sea water 



B 



.b - 


























.4 - 


















- 


SW 


70 


80 


90 


110 


120 




SW 


70 


80 


90 


110 


120 


.2 - 




























n - 





























FIGURE 2. A. Relative change in oxygen uptake of sea urchins placed in various concen- 
trations of sea water. The changes in respiration at various concentrations of sea water were 
studied in the order indicated, starting with each sea urchin in normal sea water. The initial 
oxygen consumption recorded is characteristic for each sea urchin, depending upon its size and 
other characteristics. B and C. Histograms of changes in oxygen consumption of two sea 
urchins with changes in concentration of sea water within the range of tolerance. Similar 
results (omitted here) were obtained with three other specimens. In all cases, slight changes 
in oxygen uptake were observed with variations in concentration of the bathing sea water. 
The initial oxygen uptake of each specimen depends upon its size and other characteristics. 
Comparisons should be made for the same sea urchin at different concentrations of bathing 
sea water. 

of the urchins in sea water of diverse tonicities. In each test the capillaries used 
for the melting point tests were partially filled with either the bathing fluid or the 
perivisceral fluid. The latter was obtained by inserting the capillary directly 
through the peristomial membrane into the main coelom. Care was taken not to 
contaminate such samples with the bathing fluid. The results in Figure IB show 
that the osmotic pressures of the perivisceral fluid and external media are about 
the same once equilibrium is established, i.e., no change in weight is observed. The 
differences observed at 75% and 125% sea water are probably due to incomplete 
equilibrium. 

Animals which regulate the salt concentration of their body fluids expend 



OSMOTIC RESISTANCE OF A SEA URCHIN 187 

metabolic energy and often increase their metabolic rate under conditions of osmotic 
stress (Prosser and Brown, 1961). The respiration of sea urchins exposed to 
various salinities was measured, to determine whether changes in salinity altered 
the respiratory rate. The respiratory rate for each individual was first measured 
in normal sea water. This water was then aspirated and replaced by the test solu- 
tion, and determinations were started only after 40 minutes of equilibration. In 
longer experiments one or more controls were used to safeguard against possible 
changes due to internal clocks or other uncontrollable factors. The values obtained 
for experimental animals were then corrected by the changes in these controls (for 
methods, see Farmanfarmaian, 1959). 

The data shown in Figure 2A for three series of experiments indicate that there 
is no marked change in respiration even when the urchins are immersed in 50% 
sea water, and only a small decrease when they are in 25% sea water, and most 
marked in 400% sea water. 

Within the tolerance range (70% to 120% sea water) the variation in respira- 
tion is indeed very slight as shown by 5 series of experiments. Only two of these 
experiments are shown in Figures 2B and C. 

To determine whether the osmotic stress of diluted sea water could be relieved 
by addition to the medium of an inert, non-penetrating organic compound, sea 
urchins were subjected to 50%, 25% and 5% sea water made equivalent to sea 
water in tonicity by the addition of one molal sucrose (A. Klein and R. Rasmussen, 
unpublished data), much in the manner of Loeb's study on a crustacean in 1903. 
After two hours the sea urchins appeared dead by the response criteria (see Meth- 
ods). Tests of the body fluid chlorinity by silver nitrate titration (using silver 
chromate as indicator) showed some loss of salts from the body fluid. Tests for 
reducing sugar in the body fluid (after hydrolysis with HG) were negative, indi- 
cating no marked entry of sucrose (the method would not have detected slight 
entry). Sea urchins, after two hours in each of these media, replaced in aerated 
sea water responded and appeared normal after 24 hours. At this time body fluid 
osmotic pressure (method of Gross, 1954) was normal and the weight, which had 
initially dropped, was essentially back to normal (slightly greater). When sea 
urchins were placed directly in one molal sucrose, irreversible damage occurred 
within two hours. Tests now disclosed more rapid loss of salts and marked entry 
of sucrose, suggesting that the cells lining the sea urchin membranes had died, 
and that equilibrium was being established between the body fluid and the external 
medium. 

The experiments do not tell what is happening to the cells of sea urchins placed 
in diluted media. Perhaps salt losses in them are much quicker and more marked 
than can be detected in the massive body fluid which was tested, and loss of excit- 
ability may well follow such cellular depletion. In any case, sucrose protects against 
dilution of sea water to only a limited degree, indicating that the problems which 
the sea urchin meets in media of altered osmotic pressure are ionic as well as 
osmotic. 

2. Osmotic tolerance of sea urchin eggs and developmental stages 

The results of tonicity changes on the development of sea urchin eggs are shown 
in Table II. Eggs immersed in 200% sea water perceptibly crinkled (due to loss 



188 



A. C. GIESE AND A. FARMANFARMAIAN 



TABLE II 
Osmotic tolerance of developmental stages of Strongylocentrotus purpuratus 



Per cent 
sea water 


Fertilization 
membrane 


Cleavage* to 
2 cells 


Cleavage* to 

4 cells 


Blastulae* 


Gastrulae* 


50 + 
0.5 M sucrose 


tight 


normal 


normal 


almost like control 


almost like control 


60 


about half; eggs 

swell 


occasional, ab- 
normal, delayed 


abnormal delayed 


abnormal, multi- 
cellular masses 





70 


all eggs** swell 


much like 80% 
but more exag- 
gerated 


delayed 


much delayed 
and abnormal 


" 


80 


all 


much like controls, 
blastomeres 
rounded 


delayed 


delayed, but good 


abnormal frag- 
menting 


90 


all 


much like controls 


seemingly normal 


like control 


like control 


100 


all 


normal 


normal 


normal 


normal 


110 


all 


most cleave 


seemingly normal 


like control 


like control 


120 


all 


most cleave, 
delayed 


delayed 


delayed, small 


many quite ab- 
normal 


150 


about a fifth; 
sperms active 


a few cleave 


delayed 


some blastulae*** 





100 + 
0.5 M sucrose 


sperms immo- 
bilized 


some cleave 


delayed 


abnormal 





200 


none ; sperms 
immobilized; 
eggs shrink 











* The time table for cleavage of eggs of 5. purpuralus at 13 C. has been given elsewhere (Giese, 1938). The first 
cleavage requires almost two hours at 13 C.. the subsequent cleavages occur at about hourly intervals. Free-swimming 
blastulae are found 20 hours after insemination, gastrulation occurs in about 32 hours after insemination, and the gut 
begins to differentiate in 46 hours after insemination. Plutei form in about 96 hours, although the arms do not appear 
until about 120 hours after insemination. 

** "Like controls" usually a few eggs in both control and experimental series do not develop fertilization mem- 
brane and fail to develop. 

*** In one series some abnormal blastulae developed; in one, only abnormal masses of nonmotile cells. 

of water), the sperms were immediately immobilized, and failed to penetrate the 
eggs. Later examination showed no signs of fertilization or cleavage. In 150% 
sea water the sperms were highly motile and about 20% of the eggs mixed with 
sperms developed fertilization membranes, but such cleavages as occurred were 
delayed and quite abnormal. In 120% sea water the eggs fertilized to the extent 
that controls did, but cleavages were delayed and resulted mostly in abnormal balls 
of cells. However, 24 hours later some blastulae became motile. Eggs in 60% 
sea water were also so damaged that only about half of them showed fertilization 
membranes, and only about 20% cleaved; none formed blastulae. In 70% sea 
water early cleavage was observed but later cleavages were much delayed and 
abnormal. The blastomeres of cleaving eggs were more rounded and larger than 
the controls in sea water. Eggs at the other concentrations (80%, 90%, 100% 
and 110% sea water) developed normally to the early blastula stage, and 24 hours 
later all except those in 80% sea water formed swimming blastulae. The majority 
of the embryos in 80% sea water were definitely unhealthy although a few normal 
free-swimming blastulae appeared at the same time as in the controls. Those in 
90% sea water appeared even healthier than controls in sea water. 

Twenty-four hours later, gastrulae appeared in the 110%, 100% and 90% sea 



OSMOTIC RESISTANCE OF A SEA URCHIN 189 

water; in one of 8 trials with 80% sea water, gastrulae appeared but they were 
not normal. 

It is interesting, however, that hlastulae and gastrulae transferred from the 
control in sea water to 70%, 80%, 120% and 150% sea water were alive and 
healthy 8 hours later. Those in 150% sea water were smaller than controls, and 
each had a shorter gut and a dense mass of cells around it. All survived for a week, 
at which time the experiment was terminated. At this time controls were in the 
late prism stage with skeletal rods and well-developed gut just preceding formation 
of the pluteus. Only the embryos in 110% sea water showed a degree of develop- 
ment comparable to the controls. Those in 90% sea water were almost as well 
developed but neither the skeletal rods nor the gut were comparable to those in the 
controls. Those in concentrations far to either side of sea water remained as 
enlarged and undifferentiated gastrulae, although in some of them a mouth opening 
appeared. Thus, it is apparent that not only the early development, but also the 
later development of the embryos is adversely affected by tonicities of sea water 
much removed from the normal medium. 

Changes in the development of sea urchin eggs observed in the hypotonic and 
hypertonic sea water discussed above are apparently due to changes in osmotic 
pressure rather than the quantity of the salts. This proposition was borne out by 
a series of experiments in which 50% sea water was made isosmotic with sea water 
by addition of approximately 0.5 molal sucrose (see Loeb, 1908; Harvey, 1956). 
The various stages of development from fertilization to the advanced gastrula 
are essentially normal in such a solution (see Table II). The rate of development 
and the percentage of zygotes attaining the advanced gastrula stage also are com- 
parable to controls. It must be stated, however, that embryos developing in such 
a solution appear to be a little more compact than the controls. Also, when control 
blastulae are transferred to this sucrose-sea water medium they lose water at first 
but soon attain equilibrium and appear normal thereafter. 

There is a limit, however, to which sucrose may be substituted for the salts of 
sea water without altering development. Thus, when 5% sea water is made isos- 
motic by the addition of sucrose, sperms are quickly immobilized and the fertiliza- 
tion rate drops drastically. It is known that changes in the solute environment 
alter the permeability of cells (Lucke, 1940). Development of the few eggs which 
are fertilized is completely abnormal. Also normal blastulae transferred to such 
a solution became abnormal within 6 hours. The minimal salt requirements for 
normal development of vS". purpuratus are being studied (Deboyd Smith, unpub- 
lished). Extensive studies of this type were made by Herbst (1903) on eggs of 
European sea urchins. 

The effects of a 0.5 molal sucrose solution made up in sea water were similar 
to those observed in 150% sea water. The sperms were quickly immobilized, and 
only a small percentage of the eggs were fertilized. The zygotes formed very com- 
pact abnormal blastulae, and normal blastulae transferred to the hypertonic solu- 
tion swam actively for 12 hours but became quite abnormal. 

DISCUSSION 

Echinoderms are typically marine animals but a few are found in brackish 
water, for example, Asterias rub ens, which occurs in the Baltic Sea (Schlieper,. 



A. C. GIESE AND A. FARMANFARMAIAN 

1956). This sea-star may live in a salinity as low as 8 parts per thousand (8%o) 
in the middle Baltic near Riigen. It can be moved from S%c to higher salinities 
without damage, and from higher salinities to lower ones but only gradually- 
permitting progressive adaptation over a period of several weeks. While the 
sea-stars tolerate such lowered salinities, it is interesting to note that they grow to 
a smaller size and show changes in many characteristics. For example, in water 
at I5%c as compared to 30% o righting reactions are slower, the gonads develop 
more slowly (although to the same extent), the tissue metabolism is decreased, 
and the body consists of more water and less ash. Since at lowered salinities, 
Asterias reaches osmotic equilibrium with the bathing fluid, it is possible that 
the reduction in internal salt concentration affects the activity of the enzymes in 
the cells (Schlieper, 1956). There may also be a dilution of the enzymes since 
the ratio of water to ash increases at lower salinities. 

In a study of the fauna of an estuary along the coast of Maine, Topping and 
Fuller (1942) report that Strongylocentrotus drobachiensis was found only where 
the salinity was just slightly less than in the sea. There appear to be no records 
of echinoids as resistant to low salinities as the asteroid A. rubens. 

S. purpiiratits on the California coast is probably exposed to air 3 in shallow 
pools for two- to three-hour periods a few times a year, mainly at the low tides of 
winter and summer. The exposure period is too short to allow sufficient evapo- 
ration to make the tide-pool water hypertonic. However, exposure to rain at low 
tides during the winter and spring months (December to April) may result in 
significant dilution of the sea water. 

Dilution is not general, since the local shore water tested over the year varies 
only slightly (28.7%o to 34.1S% at the Hopkins Marine Station, according to 
Feder, 1956). Considerable dilutions do occur, however, when torrents of land 
wash pour into isolated sea urchin tide-pools at low tides during and after heavy 
rainfall. On December 23, 1955, records indicate 2.7 inches of rain within 24 hours 
at Carmel, California. During the same period there were low tides of 0.0 and 
minus 0.7 foot magnitude. The sea urchins were exposed to nearly 50% sea water 
for as long as two to three hours. On February 24, 1958, titrations from similar 
pools after a rainfall of 0.72 inch (low tide of plus 0.7 foot) indicated a dilution 
to 72% sea water. It is therefore not surprising that the purple sea urchin is 
able to survive exposure to variations in the tonicity of sea water and is capable 
of withstanding, for a few hours, considerable dilution of the surrounding sea water. 
In nature, exposure to diluted sea water is probably always brief, since mixture with 
the main sea water mass quickly restores a concentration close to normal. 

Because the purple sea urchin spawns on the California coast during the rainy 
season, legend has associated spawning with dilution of the sea water, although 
no proof of a causative relationship between the two is available. Spawning, how- 
ever, does occur during this season and the eggs are fertilized and may develop 
for periods of time in diluted sea water. Since a small degree of dilution of the 
sea water does not affect development of the eggs, the rains probably do not criti- 
cally affect survival of the embryos developed in nature during this time. 

It might be argued that the resistance of developmental stages of S. purpuratus 

3 A. Klein and R. Rasmussen (unpublished) found an appreciable loss in weight and an 
increase in osmotic pressure of the body fluid of sea urchins exposed to air for 24 to 48 hours. 



OSMOTIC RESISTANCE OF A SEA URCHIN 191 

to dilution (or concentration) of sea water represents an adaptation enabling the 
sea urchin to survive the changes in tonicity it meets in its environment. A similar 
resistance of the eggs of the echiuroid worm, Urechis canpo, to variations of con- 
centration of sea water, comparable to what may occur in its native habitat, has 
also been recorded (Giese, 1954) and might also be considered an adaptation per- 
mitting survival. However, some unpublished studies with the eggs of the deep 
sea urchin, Alloccntrotns jragilis, indicate that development in this species is about 
as resistant to dilution or concentration of the sea water as is S. piirpuratus. Thus, 
the delay in cleavage of A. jragilis eggs caused by variation in concentration of sea 
water is about the same as that described above for *$". fitrpnratus, and delayed 
blastulae developed at 150%, 120%, 80% and 70% sea water, becoming progres- 
sively fewer in number and more abnormal the greater the deviation from sea water. 
Normal blastulae, comparable to controls in 100% sea water, were obtained in 90% 
and 110% sea water. Yet A. jragilis lives in deep water (Boolootian et a/., 1959) 
where appreciable changes in salinity are not recorded (Sverdrup et al., 1942). 
It must be remembered, however, that the larvae of this organism are pelagic 
(Moore, 1959) and may possibly be exposed to significant variations in salinities 
since they breed in winter (Giese, 1961 ) when rains are heavy. However, toler- 
ance, by gametes and other cells of the purple sea urchin, of changes in the osmo- 
lality of the medium is perhaps only a measure of the general tolerance of such 
changes by cells of most marine organisms. 

SUMMARY 

1. The west coast purple sea urchin, Strongylocentrotits pnrpiiratns, was found 
to be resistant to dilutions and concentrations of sea water for a brief time (three 
hours) within the range 70% to 120% sea water. 

2. The sea urchins resist exposure for many days (35) in 80% to 110% sea 
water, inclusive. Damage at higher and lower concentrations of sea water is indi- 
cated by loss of activity, loss of pigment and appendages, and failure to respond to 
food, probing, and light. 

3. The sea urchins lose weight in hypertonic solutions and gain weight in 
hypotonic solutions, presumably losing and gaining water, respectively. They 
recover nearly normal weight after replacement in sea water. 

4. The respiration of the sea urchins is little altered by changes in tonicity of 
the sea water, except at extremes of tonicity where the metabolic rate is decreased. 

5. Sea urchin eggs develop normally over the range 90% to 110% sea water, 
and considerable development occurs over the range 70% to 150% sea water. 

6. Gastrulae placed into various concentrations of sea water tolerate the change 
no better than developing eggs; differentiation proceeds in 90% to 110% sea 
water, but is essentially stopped at higher or lower concentrations even though the 
gastrulae survive. 

7. Addition of sucrose to 50% sea water to make it to up to 0.5 molal resulted 
in a medium as favorable as sea water to early cleavage and early development of 
the sea urchin egg. Addition of sucrose to sea water to make it up to 0.5 molal 
was as deleterious as 150% sea water to cleavage and early development of the 
.sea urchin egg. 



A. C. GIESE AND A. FARMAXFARMAIAN 

8. Relating the results obtained with sea urchin embryos and adults, it is 
apparent that the sea urchin is capable of withstanding the maximal variations in 
the salinity of its natural environment. 

LITERATURE CITED 

BOOLOOTIAN, R. A., A. C. GIESE, J. S. TUCKER AND A. FARMANFARMAIAN, 1959. A contribu- 
tion to the biology of a deep sea echinoid, Allocentrotus fragilis (Jackson). Biol. Bull., 

116: 362-372. 
FARMANFARMAIAN, A., 1959. The respiratory surface of the sea urchin, Strongylocentrotus 

purpuratHS. Ph.D. Thesis, Stanford University. 
FEDER, H., 1956. Natural history studies on the starfish Pisaster ochracens (Brandt, 1835) 

in the Monterey Bay area. Ph.D. Thesis, Stanford University. 
GIESE, A. C., 1938. The effect of ultraviolet radiation of X 2537A upon cleavage of sea urchin 

eggs. Biol. Bull.. 74: 330-341. 

GIESE, A. C., 1954. Osmotic tolerance of the developing Urechis egg. Anat. Rec., 120: 768. 
GIESE, A. C., 1961. Further studies on Allocentrotus fragilis, a deep sea echinoid. Biol. Bull., 

121 : 141-150. 
GROSS, W. J., 1954. Osmotic responses in the sipunculid, Dendrostoimtin sostericolum. J. 

Exp. Biol., 31: 402-423. 
HARVEY, E. B., 1956. The American Arbacia and Other Sea Urchins. Princeton University 

Press, Princeton, N. J. 
HERBST, C., 1903. Ueber die zur Entwicklung der Seeigellarven notwendigen anorganischen 

Stoffe, ihre Rolle und ihre Vertretbarkeit. III. Theil. Die Rolle der anorganischen 

Stoffe. Arch. f. Entiv., 17 : 306-520. 
LOEB, J., 1903. On the relative toxicity of distilled water, sugar .solutions and solutions of the 

various constituents of sea water for marine animals. Publ. Univ. Calif. Physiol., 

1: 55-69. 
LOEB, J., 1908. Qu'est-ce qu'une solution de saccharose isotonique pour les oeufs Strong yh- 

centrotus? C. R.. Acad. Sci. Paris, 146: 246-249. 
LUCRE, B., 1940. The living cell as an osmotic system and its permeability to water. Cold 

Spring Harbor Symp., 8 : 123-132. 
MOORE, A. R., 1959. On the embryonic development of the sea urchin Allocentrotus fragilis. 

Biol. Bull., 117 : 492-496. 
PROSSER, C. L., AND F. A. BROWN, JR., 1961. Comparative Animal Physiology. Second ed. 

Saunders, Philadelphia. 
SCHLIEPER, C., 1956. Comparative study of Asterias rubcns and Mytilus cdulis from the North 

Sea and the western Baltic Sea. Ann. Biol., 33: 117-127. 
SVERDRUP, H. V., M. JOHNSON AND R. FLEMING, 1942. The Oceans. Prentice Hall Inc., 

New York. 

TOPPING, F., AND J. FULLER, 1942. The accommodation of some marine invertebrates to re- 
duced osmotic pressures. Biol. Bull., 82 : 372-384. 



FACTORS OF EQUATORIAL CONTRACTION AND POLAR 
MEMBRANE EXPANSION IN ENDOSPERM CYTOKINESIS 

HENRY S. ROBERTS 
Department of Zoology, Duke University, Durham, North Carolina 

In a recent review (Roberts, 1961) I suggested that cytokinesis in both plant 
and animal cells might be explained in terms of the factors of polar expansion, 
equatorial contraction, and new membrane formation. Mazia (1961) has also 
discussed basic similarities in cytokinesis of plant and animal cells. There is con- 
siderable support for the idea that new membrane formation may occur in the 
connecting stalk of dividing animal cells late in cleavage (references and discussion 
in Roberts, 1961; Mazia, 1961; Buck and Tisdale, 1962a, 1962b). Buck and 
Tisdale (1962a) have demonstrated in rat erythroblasts the formation of a mid- 
body in the equator of the spindle during cleavage, which appears to correspond 
in structure to the phragmoplast of plant cells, and (1962b) have described in 
three types of mammalian cells membrane-bound vesicles, apparently derived from 
endoplasmic reticulum, which develop in the cleavage plane and apparently are 
the source of new cell membrane in the furrow region. Their observations are 
somewhat similar to those of Porter and Machado (1960) on cell-plate formation 
in Allium. However, there has been almost no evidence that polar expansion and 
equatorial contraction might play a part in cytokinesis of plant cells which employ 
cell-plate formation, other than the observations of Bajer and Mole-Bajer (1956) 
on the division of well flattened living endosperm cells. "The lack of cellulose 
walls makes possible the formation of cytoplasmic protuberances. These pseudo- 
podia-like strings of cytoplasm may form during the whole division, but are most 
prominent in prophase and tclophase. During pro- and metaphase the shape of the 
cell often changes to a sphere or an ellipsoid often without pseudopodia formation. 
It is of interest that in other material the cells often become round before meta- 
phase. In animal tissue it is well known that pseudopodia disappear before ana- 
phase. . . . Endosperm cells often form out pushing s during the tclophase." These 
suggestions of surface forces in dividing endosperm, similar to those of dividing 
animal cells, stimulated the writer to look at dividing living endosperm cells which 
had not been distorted by flattening. 

Investigators are rightfully concerned with standards of normalcy in studies of 
living cells, and with conclusions based on cells studied under abnormal conditions. 
Some would reject all studies of cell division based on cells which do not complete 
at least two divisions without evidence of abnormality. Such a criterion would 
impose serious and frequently impossible difficulties on the study of terminal 
divisions, as in secondary spermatocytes, and eliminate as subjects of investigation 
many cells with long division cycles. Other investigators recognize that degrees 
of abnormality are the almost inevitable result of preparation and observation. 
They believe that there is information to be gained from abnormal cells, provided 

193 



194 HENRY S. ROBERTS 

the investigator recognizes and makes extremely clear the limits of his methods 
and of his results, and is extremely cautious in his conclusions. The observations 
here reported were made on cells which do not meet the criterion above and which 
were dividing under abnormal conditions. The limitations are detailed in the 
body of the paper and are considered in the discussion. Subject to these limitations 
the observations support the idea that equatorial contraction and polar expansion 
may play a significant role in cytokinesis of endosperm cells lacking a cell wall. 

MATERIALS AND METHODS 

Young seeds still in the milk stage of whatever horticultural varieties of 
Hemerocallis were available at the moment, were the source of endosperm. Seeds 
collected at mid-day or early afternoon provided the best material. Those collected 
in the early morning contained few dividing cells. One end of the seed was sliced 
off with a sharp scalpel, and endosperm was sucked out of the cavity with a 1-ml. 
syringe fitted with a 24-gauge needle whose tip had been ground to roundness. 
The seeds were not squeezed to express more material. Four to eight seeds pro- 
vided enough material for a preparation. A 5-7-mm. square was outlined on a 
no. 1 coverslip with Vaseline extruded through a 24-gauge hypodermic needle, 
and the enclosed space flooded with endosperm fluid. Excess fluid was carefully 
withdrawn with the syringe, leaving a relatively flat film whose depth could be 
adjusted with some experience. The coverslip was then inverted over the shallow 
depression of a Fisher-Littman well slide and sealed by a ring of Vaseline. Prepa- 
rations were examined and photographed by phase contrast microscopy using a 
12.5 X compensating ocular and 45 X dark contrast objective. The microscope 
(Spencer) was equipped with long focal length phase plates. Illumination was 
provided by a ribbon filament lamp with 36 mm. of 3% copper sulfate solution in 
the light path as a heat absorber. Light intensity was controlled by a variable 
transformer and was kept at the minimum possible for observation except for the 
few seconds required for photographic exposure. 

The preparations differ significantly from those of the Bajers (1954; 1956). 
Thick hanging drop preparations were necessary to avoid flattening. This intro- 
duced problems already described by Bajer (1954, p. 386). ". . the thickness of 
the drops is of greatest importance. In large drops the penetration of oxygen is 
not sufficient, and in consequence most cells die in prophase, though most mitoses 
in metakinesis are continued to t el o phase." Bajer found it necessary to use drops 
only six micra in thickness. Since freely suspended cells were essential a soft agar 
substrate could not be used. Most of the first season of study was devoted to efforts 
to achieve longer-lived preparations, without significant success. It was thus neces- 
sary to accept the limitations imposed by the requirement for thick preparations 
and freely suspended cells, and resort to the simple preparations described. Al- 
though the small air space of the thin slides used was rapidly saturated, as judged 
by condensation, evaporation ensured that the medium was at least mildly hyper- 
tonic. Preparations thus made were short-lived. After H-2 hours the cells showed 
obvious signs of abnormality and division soon ceased. Cells first observed in 
prophase could be followed to metaphase. Those first observed in metaphase 
could be followed to late anaphase. Cells in middle or late anaphase could be fol- 
lowed to telophase and completion of the cell plate. 



CYTOKINESIS IN ENDOSPERM 



195 



Routinely each preparation was completely scanned by overlapping traverses 
and the location of suitable cells for further observation and photography noted. 
Such scanning was completed in approximately five minutes. It is emphasized that 
all of the stages described were regularly observed during the preliminary scan- 
ning, long before there were evidences of abnormality. The investigation extended 
through tw r o summers. Seventy cells were followed in detail, and 13 divisions 
recorded photographically. Incidental observations and photographs were made 
of many more. Only those cells which continued to divide, subject to the limita- 
tions previously described, are the basis of the following observations. 

OBSERVATIONS AND DISCUSSION 

The preparations contained abundant participate material, ranging in size from 
barely visible to 1-2 micra, cellular debris, bits of peripheral endosperm tissue 
with well developed cell walls, and free nuclei. The numerous small particles are 
a source of difficulty in observing and photographing cells deep in the fluid but 





B 





H 





M 



N 










K 





R 





u 



w 



X 



FIGURE 1. Comparison of cytokinesis of animal cells and endosperm. A-F, cytokinesis 
of animal cells. In C-D polar membrane expansions, which may take different forms or may 
be demonstrable only indirectly in various cell types, are shown as surface irregularities. 
G-L, cytokinesis in endosperm. Equatorial constriction and the two types of surface irregu- 
larities are shown in I-J. M-R, failure of cell plate formation and of cytokinesis in endosperm, 
based on a single cell. S-X, absence of equatorial constriction and polar irregularities followed 
by failure of cytokinesis. 



196 



HENRY S. ROBERTS 




FIGURES 2-16. 



CYTOKINESIS IN ENDOSPERM 197 

the presence of the smallest particles in active Brownian movement above the cells 
being observed ensured that they were not compressed between the coverslip and 
the surface film. The free nuclei were not observed to divide nor were recognizable 
prophases of division seen. It is worthy of comment that most of these nuclei 
had adherent granular material which could not be optically distinguished from 
the smaller particles seen in the endosperm fluid but lacked visible cell membranes. 
Division was observed only in cells with cytoplasm and enclosed in a cell mem- 
brane. The cells, when flattened, were entirely comparable to those studied by the 
Bajers (1954, 1956) and are believed to be of the same type. Their origin within 
the endosperm sac is unknown but is presumed to be the same gelatinous layer 
which the Bajers expressed from the seed. 

Figure 1 summarizes the observations on the process of cytokinesis and pro- 
vides a comparison with similar stages in animal cells. At metaphase the cells 
round up (Figs. 1G, 2), much as described in the previous quotation from Bajer. 
During anaphase the cells elongate (Fig. 1H), and in late anaphase or early telo- 
phase develop a distinct equatorial constriction (Figs. 1, I-J, 3-5), entirely 
comparable in appearance to early furrow formation in animal cells. At about the 
same time irregularities appear in the membrane at the cell poles. These irregulari- 
ties are of two types ; small bulges which appear in the same location and at the same 
stage of division as the blebs and visible expansions characteristic of cleaving animal 
cells, and long oriented spinous processes (Figs. 1 I-J, 3-6). Small particles, which 
have become increasingly visible in the spindle, aggregate in the equator and fuse 
to form the cell-plate which extends peripherally, eliminating the equatorial con- 
striction and producing a pronounced bulge at the equator (Figs. IK, 7-8). The 
polar membrane irregularities gradually disappear, leaving a configuration like 
that seen in Figures IK. 7-8. This gradually shortens to the condition seen in 
Figures 1L, 9, the final stage which could be followed. Except that they are at all 
times enclosed in a cell membrane, the configurations assumed by the dividing cells 
are quite comparable to those shown by Jungers (1931). 

The similarity to animal cell cytokinesis is striking and is well illustrated in 
Figure 1. The manner in which such an illustration is drawn can dramatize 
similarities and de-emphasize differences, but reference to the photographs (Figs. 
2-16) should be convincing, especially to those who have observed division of 
freely suspended animal cells. The observations establish the fact of equatorial 

FIGURES 2-3. Metaphase and anaphase of the same cell. Figure 3 shows equatorial 
constriction, membrane irregularities and one spinous process. In these and the following 
figures the scale, 480 X, is shown by the small rectangles, whose length indicates 10 p. 

FIGURES 4-5. Middle and late anaphase of the same cell, showing development of equa- 
torial constriction and polar irregularities. Comparison with Figures 2-3 indicates range 
of cell size observed. 

FIGURE 6. Cell plate formation in early telophase. The spinous processes are well 
developed but not unusually so. 

FIGURES 7-9. Telophase and the completion of cytokinesis. Figure 9 is the same cell 
shown in Figure 6. 

FIGURES 10-11. Failure of cell plate formation and maximum observed equatorial con- 
striction. In this cell cytokinesis failed and a binucleate resulted. 

FIGURES 12-14. Absence of equatorial constriction and membrane irregularities. The 
cell plate in Figure 14 appeared to be degenerating. 

FIGURES 15-16. "Amphiastral" configurations in medium made hypertonic with glucose. 
The spinous processes suggest astral rays. 



198 HENRY S. ROBERTS 

constriction but of themselves tell us nothing of its nature. It does not appear 
to result from conformation of the cell membrane to the contours of the internal 
spindle and re-forming nuclei. Small particles in Brownian movement may occa- 
sionally be seen between the spindle and equatorial constriction, and around 
the re-forming nuclei. There is evidence the equatorial constriction in animal cells 
is the result of an equatorial contraction (reviewed in Roberts, 1961 j. I suggest 
that the same mechanism may be operative in these endosperm cells, probably to 
a lesser degree. The observed changes in the polar cell membranes undoubtedly 
are accompanied by an increase in membrane area at the poles. Their occurrence 
at the same stage of division as the better known polar expansions of animal cells 
suggests they may play a similar but perhaps less important role in cytokinesis of 
endosperm cells. The long oriented spinous polar processes deserve special com- 
ment. They are straight, or nearly so, and if the long axes are projected back- 
ward, all converge in the re-forming nucleus. Occasionally in routine preparations 
and more commonly in preparations made more hypertonic by the addition of 
glucose (Figs. 15-16), the spines develop so abundantly that the cells resemble 
diagrammatic amphiastral figures. Wilson ( 1900 ) described somewhat similar 
''filose" processes extending from the poles of the elongated polar bodies and early 
blastomeres of Cerebratulus. The centers of "filose" activity were correlated with 
the position of the spindle poles of the amphiastral spindles. Lima-de-Faria (1958), 
Ostergren (1954), and Ostergren, Koopmans and Reitalu (1953) have described 
astral rays in numerous genera and species of higher plants. They are in agree- 
ment that the astral rays are small at metaphase and reach maximum development 
at late anaphase. Appearance and time of development suggest that the observed 
spinous processes of endosperm cells may be associated with astral development. 
However, astral rays have not been observed by the writer and other explanations 
are possible. Occasionally cells have been observed to divide with only minor polar 
membrane irregularities, without the development of spinous processes. 

In a single instance a cell (Figs. 1 M-R, 10-11) was observed to progress to 
late anaphase quite normally and then fail to form a cell plate. During early 
telophase the constriction slowly deepened, advancing further than usual but falling 
far short of dividing the cell into two (Fig. 11). During telophase the cell sud- 
denly rounded up, resulting in a binucleate. More frequently cells, identifiable in 
early anaphase by their distinctive outlines (Figs. 1 S-X, 12-14), failed to develop 
either irregularities of the polar membrane or equatorial constrictions. A cell 
plate formed quite normally but in all cases failed to contact the equatorial cell 
membranes. In telophase the cells rounded up and the cell plate showed signs of 
degeneration. Presumably they too would have formed binucleates, but the life 
of the preparations was too short to observe the final result. Occasionally bi- 
nucleate cells were observed during the first minutes of a fresh preparation. It may 
be coincidental but is probably of significance that in every case in which one of 
the factors of equatorial constriction, polar membrane expansion, or cell plate 
formation failed, cytokinesis failed. 

Observations such as those reported cannot of themselves establish that equa- 
torial contraction and polar membrane expansions occur and are factors in the 
mechanism of cytokinesis of normal endosperm cells. The limitations of the meth- 
ods used are such that the observations reported could be ascribed to the effects 



CYTOKINESIS IN ENDOSPERM 199 

of the methods employed. However, it is not without significance that in hundreds 
of preparations, during two seasons, every cell observed to divide did so as de- 
scribed. All of the stages described have been observed during the first few minutes 
after preparation. Further, the observations reported were made during the 14-2 
hours before signs of deterioration appeared. On this basis the observations do 
suggest a strong probability that equatorial contraction and polar membrane ex- 
pansions occur normally and are significant factors in the mechanism of cytokinesis 
of endosperm cells. 



Drs. Sally Hughes- Schrader and Lewis Anderson read the manuscript and 
made valuable comments and suggestions. I gratefully acknowledge their assistance. 

SUMMARY 

Divisions of living endosperm cells of Hemerocallis have been observed in 
hanging drop preparations where they are not distorted by flattening. Equatorial 
constriction and polar membrane expansions occur in late anaphase and appear 
quite comparable to corresponding stages of cytokinesis in animal cells. Failure 
of equatorial constriction, of polar membrane expansion, or of cell plate formation 
results in failure of cytokinesis. It is suggested that equatorial contraction and 
polar membrane expansion may be functionally important in the mechanism of 
cytokinesis of endosperm cells, as they are in animal cell cytokinesis. 

LITERATURE CITED 

BAJER, A., 1954. Cine-micrographic studies on mitosis in endosperm. I. Acta Sue. Botan. 

Poloniae, 23: 383-412. 
BAJER, A., AND J. MOLE-BAJER, 1956. Cine-micrographic studies on mitosis in endosperm. 

II. Chromosome, cytoplasmic and Brownian movements. Chromosoma, 7 : 558-607. 
BUCK, R. C, AND J. N. TISDALE, 1962a. The fine structure of the mid-body of the rat 

erythroblast. /. Cell Biol, 13: 109-115. 
BUCK, R. C., AND J. N. TISDALE, 1962b. An electron microscope study of the development 

of the cleavage furrow in mammalian cells. /. Cell Biol., 13: 117-125. 
JUNGERS, V., 1931. Figures caryocinetiques et cloisonnement du protoplasme dans I'endosperm 

d'Iris. La Cellule, 40: 293-359. 
LIMA-DE-FARIA, A., 1958. Recent advances in the study of kinetochore. Pp. 123-157 in 

International Revietv of Cytology, Vol. Ill, G. Bourne and J. F. Danielli, eds. 

Academic Press, N. Y. 
OSTERGREN, G., 1954. Astral rays in normal and chemically disturbed mitoses of higher plants. 

Congr. Intern, botan., 8 e Congr., Paris, 1954, sects. 9-10, pp. 15-16. 
OSTERGREN, G., A. KOOPMANS AND J. REITALU, 1953. The occurrence of the amphiastral type 

of mitoses in higher plants and the influence of aminopyrin on mitosis. Botan. Notiser, 

106: 417-419. 
MAZIA, D., 1961. Mitosis and the physiology of cell division. Chapt. 2, pp. 77-412 in The 

Cell, Vol. Ill, J. Brachet and A. E. Mirsky, eds., Academic Press, N. Y. 
PORTER, K. R., AND R. D. MACHADO, 1960. Studies on the endoplasmic reticulum. IV. Its 

form and distribution during mitosis in cells of onion root tip. /. Bioph\s. Biochem. 

Cytol, 7: 167-180. 
ROBERTS, H. S., 1961. Mechanisms of cytokinesis, a critical review. Quart. Rev. Biol., 36: 

155-177. 
WILSON, C. B., 1900. The habits and early development of Cerebratulus lacteus (Verrill). 

A contribution to physical morphology. Quart. J. Micr. Sci., 43 : 97-198. 



ECHOES OF ULTRASONIC PULSES FROM FLYING MOTHS 

KENNETH D. ROEDER 

Department of Biology, Tufts University, Medford 55, Massachusetts 

It has been firmly established by Griffin, Webster and Michel (1960) that 
flying bats use echolocation in tracking and capturing flying insects. By this 
means Myotis lucifugus is able to detect insects as small as Drosophila melanogaster 
at distances as great as 50 cm. These authors also used acoustic jamming 
experiments to show that under these circumstances the bat's clue to the presence 
of an insect is the echo of its ultrasonic cry rather than the sonic output from 
the vibrating wings of the prey. However, they reaffirmed the observation made 
by others (Mohres, 1950; Kolb, 1959; Treat/ 1955) that bats at rest and not 
emitting ultrasonic orientation sounds will respond quickly and accurately to the 
nearby presence of a buzzing insect. 

The ability of Myotis to determine the distance, direction, and presumably the 
size of an object as small as Drosophila at a distance of 50 cm. suggests that bats 
might be able to discriminate many other characteristics of larger prey, such as 
moths, through the properties of their echoes. Most observers of bat behavior 
have at one time or another tossed pebbles or other inert objects to feeding bats. 
In the field bats will commonly detect and track such objects, but rarely attempt 
to attack or capture them. Under laboratory conditions bats can be trained 
to track and capture non-flying objects such as mealworm larvae tossed in the 
air (Webster, 1963), but under natural conditions it seems unlikely that they 
ever encounter potential prey that is not flapping its wings. Although the wing 
sound seems to be unimportant as a clue (Griffin, Webster and Michel, 1960), it 
seems possible that the wing movement of the prey may modify the echoes 
returning to a flying bat and enable it to discriminate a flying insect from a pebble. 
This possibility was first pointed out by Griffin (1958). 

Further indirect evidence in favor of this hypothesis is the often-made observa- 
tion (e.g., Treat, 1955; Roeder, 1962) that "freezing" is one of the many types 
of response made by moths to the proximity of bats or when exposed to ultra- 
sound. Under field conditions some moths (and lacewings) cease all flight move- 
ment and drop to the ground when exposed to a source of ultrasonic pulses. In 
one sense this maneuver might be expected to simplify the bat's task of tracking 
its prey, since the tracker is presented with a target falling in a roughly ballistic 
and therefore more predictable trajectory compared with a moth's flight path. 
Since "freezing" behavior can be presumed to have some survival value to the 
moth, this disadvantageous aspect (to the moth) is possibly offset by the attain- 
ment of some degree of acoustic concealment. In passing, it should be pointed 
out that exposure to a series of bat-like sounds just as frequently causes moths 
to make a variety of violent evasive maneuvers or to fly directly away from the 
source (Roeder, 1962). It is the age-old question of whether it is best to duck. 

200 



ULTRASONIC ECHOES FROM MOTHS 



201 



dodge, or run, to which the answer seems to be that there is survival value in 
variety. 

At present there seems to be no direct way of finding out how much information. 
a bat obtains from echo fluctuations produced by the wing movements of its prey. 
In the following experiment the bat was replaced by a source of ultrasonic pulses 
and a microphone in order to determine how the amplitude and other properties 
of the echo are affected by the wing movements of flying moths. 




FIGURE 1. Block diagram of apparatus used to record the attitude and echo of a moth 
in fixed flight. Dotted line, acoustic path ; dashed line, optical path, or other details 
see text. 

METHOD 

A block diagram of the apparatus is shown in Figure 1. In the early experi- 
ments a 35-mm. camera (Exakta) equipped with a 180-mm. telephoto lens framed 
the tube face of an oscilloscope at a distance of one meter. Half the image of 
the tube face was occupied by a prism (P) at 50 cm. that served to align the 
image of the moth next to that of the tube face. Since both images were at the 
same distance from the camera, the centimeter scale on the tube face (Fig. 2 
et seq.} serves both as a time base for the echo trace and a size scale for the 
image of the moth. 



202 KENNETH D. ROEDER 

Grouped as closely as possible around the prism were the sound source (S) 
an ultrasonic transducer of the capacitative type with a membrane of 0.5 mil 
metallized Mylar, a Granath microphone (M) operating on the same principle, 
and the flash unit of a Grass PS1 Photo Stimulator (F). All three units were 
aimed as precisely as possible at a point to be occupied by the moth 50 cm. distant 
and at right angles to the camera-oscilloscope axis. The apparatus was aimed 
acoustically as well as visually by placing a 1-cm. sphere at this point and maxi- 
mizing the echo picked up by the microphone. An attempt was made to place 
the axes of the microphone and transducer in relation to the optical axis so 
that the optical profile of the moth would correspond as closely as possible to that 
presented to the sound pulse. 

In most of the experiments the ultrasonic pulses were between 70 and 90 
kcps, 0.7 millisecond in duration, and with about 0.2 millisecond rise and fall time. 
Frequency modulation, of the sort encountered in the pulses of vespertilionid bats, 
was not employed. The ultrasonic pulses were formed by clipping the output 
of an oscillator (OSC.) by means of an electronic switch (General Radio). The 
"on" (A) and "off" (B) commands to the switch were the synchronizing and 
stimulus pulses from a Grass S4 stimulator. Thus, the ultrasonic pulse duration 
was controlled by the stimulus delay circuit of the stimulator. The flash contact 
in the camera triggered the stimulator each time a frame was exposed. The ultra- 
sonic pulses thus formed passed through an attenuator and amplifier to the trans- 
mitter. In some of the experiments the signal leaving the attenuator was 
monitored on the lower beam (Y 2) of the oscilloscope. The sweep of the latter 
was triggered by the A pulse from the stimulator. A sweep speed of 1.0 
millisecond/mm, was used throughout. 

The A pulse also triggered the electronic flash after passing through a delay 
circuit. The delay was needed to insure that the image of the moth was photo- 
graphed at the instant the sound pulse reached it. The picture thus reveals the 
moth's attitude in flight at the moment the sound pulse was reflected from its 
'wings and body. Since the moth was approximately 50 cm. distant from the 
sound source the flash was delayed by 1.6 milliseconds. The duration of the 
flash was approximately 10 microseconds. 

The microphone used to detect the echo was connected through its amplifier 
and a band-pass filter (to eliminate extraneous noise) to the upper beam (Y 1) 
of the oscilloscope. 

Moths of various species were captured at light. They were mounted by 
cementing the mesonotum with Tackywax to an insect pin. The pin was attached 
to a thin vertical support of bamboo. Pin and support alone gave a negligible 
echo. No other object stood for a radius of several feet behind the moth, but 
little trouble was experienced with extraneous echoes, owing to the brevity of 
the pulses and the high sweep speed. The click of the camera shutter produced 
some acoustic interference, but this arrived later than the echo and could be 
disregarded. 

Most moths flew spontaneously, and often for considerable periods, as soon as 
a paper wad in contact with the tarsi was removed. The procedure was to take 
a sequence of frames at random as the moth continued to fly. Sixty to 70 frames 
covered most of the wing positions for a given angle of presentation. Each frame 



ULTRASONIC ECHOES FROM MOTHS 203 

was exposed for 0.01 second. The opening of the camera shutter via the cable 
release triggered the onset of the ultrasonic pulse, the sweep, and, after an appropri- 
ate delay, the flash. The lower trace in each frame displays the shape of the 
signal and the upper trace displays its echo as detected by the microphone. 

Some direct interaction took place between transmitter and microphone. This 
occurred because a fairly intense pulse was needed in order to get an adequate 
echo from the moth, and because it was necessary to place transmitter and micro- 
phone as close as possible above and below the prism so as to minimize the dis- 
parity between the optical and acoustic reflections. The signal caused by this 
interaction appears as the first pulse on the upper trace of each recording, and 
serves as another control of the outgoing signal. During a given experiment it 
remained constant. The second signal on the upper trace, occurring about 2.5 
milliseconds after the direct pulse, is the echo returned by the moth while it was 
in the attitude shown by the accompanying flash picture. 

The still camera was replaced in later experiments by a 16-mm. motion picture 
camera equipped with a telephoto lens. The rotating shutter was equipped with 
a wiping contact that closed momentarily during the exposure of each frame. 
This triggered the stimulator, sound pulse, and flash as before. The camera 
(Ensign) was a spring-driven model, about 35 years old, in which the film 
speed was found to vary with the amount of pressure applied to the release button. 
This defect made it possible to obtain satisfactory stroboscopic pictures. With the 
moth in flight the camera speed was gradually increased until the number of frames 
per second approached the wingbeat frequency (commonly between 15 and 35 per 
second) of the moth. This was determined by watching the image of the 
moth revealed by the flashes of the strobe light. The large number of frames 
made available by the motion picture was of great value in the analysis. 

In view of what has been said (Roeder, 1962) about the behavior of moths 
in the presence of ultrasound it may seem paradoxical that the moths continued 
to fly while being bombarded by high-intensity ultrasonic pulses. Indeed, at the 
beginning of a run they frequently ceased flying as soon as the sound sequence 
began. Some specimens were discarded as being too refractory or erratic. How- 
ever, continued exposure to sound appeared to adapt the neural mechanism respon- 
sible for evasive behavior, and in most cases the moths flew steadilv after a few 

j 

false starts. It is fairly certain that many of the recorded wing attitudes included 
abrupt changes in angle of attack and amplitude associated in free moths with 
erratic flight. These and similar departures from "normal" flight movements are 
also to be expected from the restrained condition of the subjects. This problem 
of studying natural wing movements in insects restrained for observation has long 
plagued students of insect aerodynamics. 

Echoes were recorded from the following species: Sunira bicolorago Gn., 
Amphipyra pyramidoides Gn., Agrotis ypsilon Rott., Enargla decolor Wlk., En- 
nouws inognai'iiis Gn., Aniathcs c-nigrum L., Graptolitha ituiinoda Lintner, and 
Orthosia Jiibisci Gn. 

RESULTS 

Figure 2 shows certain attitudes and their corresponding echoes when the 
axis of a flying moth is approximately at right angles to the sound path. At this 



204 KENNETH D. ROEDER 

angle the largest echo is produced when the wings are near the top of the stroke 
(A). The smallest echo is produced during the latter part of the downstroke (C), 
while echoes of intermediate size occur at the beginning of the downstroke (B) 
and near its end (D). The maximum echo produced near the top of the wing 
stroke is dependent upon a critical wing angle. This caused it frequently to be 
missed when single frames were taken. The stroboscopic motion picture method 
not only made a much greater number of single frames available, but it was also 
possible by manipulating the camera speed to hold a particular wing attitude for 
a number of consecutive frames. Figure 3A illustrates the contrast in magnitude 




FIGURE 2. Echo and attitude of Agrotis ypsilon in fixed flight on bearing about 90 to 
sound path. Ultrasonic pulse 78 kcps, 0.7 msec, in duration. Grid on oscillogram equals 
1.0 msec, on trace, 1.0 cm. on photograph. A, top of wing stroke; B, first half of downstroke; 
C, last half of downstroke ; D, bottom of stroke. See text for other details. 

between minimum and maximum echoes produced during the last half of the 
upstroke. In this series also the maximum echo appears to be produced when 
the surface presented by the wings is about 90 to the sound path. An attempt 
to hold the optically determined attitude of the moth constant at this point by 
adjusting the camera speed produced a series of attitudes that show little optical 
difference, although the size of the echo fluctuates widely (Fig. 3B). 

From this it can be concluded that the body of the moth plays a negligible part 
in causing an echo, most of the acoustic reflection coming from the surface of 
the wings presented at 90 to the sound path. This effect may at times be 
further accentuated by the slightly curved surface assumed by the wings at the 
beginning of the downstroke. This maximum echo must occur either once or 



ULTRASONIC ECHOES FROM MOTHS 205 

twice in rapid succession near the top of the stroke, depending upon the amplitude 
of the stroke and upon the precise angle of the moth relative to the sound path. 

Recordings made with the axis of the potential flight path at other angles 
to the sound path showed similar fluctuations of echo with wingbeat, hut the 
size of the maximum echo never reached that recorded when the axis of the 




FIGURE 3. Consecutive frames from motion picture of echo and attitude of flying Orthosia 
hihisci. Ultrasonic pulse 85 kcps, 0.7 msec, in duration. Grid equals 1.0 msec, on oscillogram, 
1.0 cm. on photograph. A, the second half of the upstroke. B, image held nearly stationary 
in stroboscopic sequence. C, path of moth directly away from sound source, wings in phases 
of upstroke. 



206 



KENNETH D. ROEDER 



moth was approximately 90 to the sound path. Samples of echoes from other 
angles are shown in Figure 3C, where the moth is headed directly away from 
the sound source, and in Figure 4, where the course is about 135. 

In the present experiments the moth was always mounted as if it were in 
level flight at the same altitude as the optical and acoustic system. Since the 
surface of the wings appears to be the main source of echoes, it is apparent from 
the photographs, particularly those of Figure 4, that maximal echoes would be 




FIGURE 4. Echo and attitude of A gratis ypsilon on course about 135 away from sound 
source. Other details as in Figure 2. A, early in downstroke. B, mid downstroke. C, early 
in upstroke. D, mid upstroke. 

returned from other wing positions and flight angles if the transmitter and micro- 
phone were aimed at the moth at various angles from above or below the 
plane of flight. These were not investigated. 

During a given run the peak echoes usually went off the oscilloscope screen 
(Fig. 3B) while minimum echoes sometimes disappeared in the noise level of 
the recording system. This made it difficult to estimate with any accuracy the 
intensity difference between maximum and minimum echo. In many cases it 
was certainly greater than 20 to 30 decibels. 

Distortions of the echo were also common. The echo shown in Figure 5A 
has essentially the same form as the outgoing pulse. The asymmetric peak in B 
was probably due to movement of the wings towards the attitude producing a 
maximum echo during the interval of time (0.7 millisecond) that the pulse 
impinged upon them. The sharp peak shown in C indicates that the attitude 



ULTRASONIC ECHOES FROM MOTHS 



207 



producing a maximal echo was reached only briefly near the midpoint of the pulse. 
The double peaks shown in D and E may have been due to the opposite effect, 
the wings passing through an attitude of minimum echo during the pulse. Another 
possible explanation of the double peaks shown in D and E is suggested by the 
attitudes of the wings shown in the photographs. Echoes may have been 
produced separately by the near and far pair of wings. A difference in the 
length of the sound path from one of the two reflecting surfaces to the source 
by 0.5 or 1.5 wave-lengths might be expected to produce partial or complete 
interference and extinction of the echo. For an 82 kcps pulse of 4 mm. wave- 
length, such interference w r ould occur when the wings were 1.0 or 3.0 mm. apart. 
This cannot be measured from the photographs, but the dimensions of the moth 
make it entirely possible. 




FIGURE 5. Distortions in the echo returned by Orthosia hibisci. Other details as in 
Figure 3. A, echo showing minimal distortion. B, distortion due to movement of wings to 
point of greater reflectance during impact of pulse. C, sharp peak caused by wings moving 
through position of maximum reflectance during pulse. D and E, double echoes ; see text. 
F, abnormally small echo. 

A similar explanation may account for cases where the minimum echo is 
below the noise level or possibly absent (Fig. 5F). Occasionally, as in D and E, 
a small pulse appears somewhat more than a millisecond later than the main 
echo. The origin of this is unknown. 

Some speculation has centered on the functional significance of the scales that 
typically cover the wings and body of Lepidoptera. The setae covering the 
thorax may be extremely dense and filiform, forming a deep "fur" in many 
noctuids, such as Lencania pseudargyria. In other species, particularly in certain 
arctiids, this coat of scales may be verv thin or almost absent, yet moths of both 

J J - J 

families and with all degrees of thoracic vestiture have well-developed tympanic 
organs (Haskell and Belton, 1956; Roeder and Treat, 1957) and are presumably 
subject to attack by bats. 



208 KEXXETH D. ROEDER 

A few measurements were made of the effect of this covering of scales on 
the echoic qualities of the subjects of the present study. Amputated wings were 
statically mounted so as to produce a maximum echo. They were then denuded 
of scales with an artist's paintbrush and the echo re-measured. Similar treatment 
was given to wingless bodies. Removal of the scales increased the echo by 
1 or 2 decibels. Since this is an insignificant figure compared with that 
produced by changes in wing angles during normal flight, it was concluded that 
the scales play an unimportant role in reducing the echoic qualities of moths. 

DISCUSSION 

From one point of view it might seem that these results demonstrate merely 
what could have been predicted from an elementary knowledge of the laws of 
wave motion. There is some comfort in this to the biologist accustomed to the 
unexpected in living things. The size of the echo from the wings in the attitude 
normal to the sound path compared with that produced by the body and the 
wings in other attitudes shows that the acoustic profile of a flying moth goes 
through much greater extremes than does its optical profile. Assuming that a 
bat were equipped only with a crude sonar system of the sort used in these 
experiments, the most definitive information that it would receive would be that 
its prey was flapping its wings. An optical comparison is suggested by the 
scintillations produced by suspended microscopic crystals, such as mica. How- 
ever, in the biological situation occupied by the hypothetical bat and a flying moth, 
the scintillations would occur at regular intervals determined by the wingbeat 
frequency of the insect, and they would be maximal only when their source was 
travelling on certain bearings relative to the flight path of the bat. 

It is certain that this postulated situation is a gross over-simplification, 
particularly with respect to the acoustic capabilities of the bat. Nevertheless, it 
does suggest some points that may have relevance in connection with the 
behavior of flying moths when exposed in the field to a sudden train of ultrasonic 
pulses (Roeder, 1962). When close to the ultrasonic source or when exposed 
to pulses of high intensity many moths react by changing from a relatively 
straight flight path to a variety of turns, spirals, and dives. Others close their 
wings and fall passively to the ground. The experiments reported here suggest 
that cessation of flight movements and closure of the wings must eliminate the 
major source of echoes, as well as the echo fluctuation characteristic of flight, 
thereby providing the insect with some measure of acoustic concealment as it 
falls to the ground. 

In the same paper it was reported that moths flying at greater distances from 
the sound source and exposed to lower intensities frequently turned from their 
flight path and flew directly away from the sound source. This maneuver has the 
obvious advantage to the moth in putting distance between it and its potential 
predator, but the echo experiments suggest that it may have additional survival 
value. Moths flying at roughly the same altitude as an approaching bat are 
most likely to present an optimum target for echoes if they cross the flight patli 
of the bat at about 90 (see Fig. 3). The first clue available to a bat approaching 
at maximum range must be a very brief echo of its cry occurring once or twice 
for each wingbeat of the target. Since at least some noctuid moths are certainly 



ULTRASONIC ECHOES FROM MOTHS 209 

capable of detecting the echolocating cries of a bat at a considerably greater range 
than a bat can detect their echoes (Roeder and Treat, 1961), it must be of some 
advantage to the moth to assume a less echo-producing flight path, e.g., parallel to 
or directly away from that of the bat, as soon as the latter has been detected. 

The situation is much more difficult to assess from the viewpoint of the bat, 
for its frequency-modulated cry is more complex than the pulses used in these 
experiments and little is known about its capabilities of acoustic discrimination. 
Relatively long pulses, such as the cruising pulses emitted by Myotis liicifugus 
(Griffin, 1958), might increase the bat's chances of picking up a. brief maximal 
echo from a flying moth. For instance, a cry 15 milliseconds in duration would 
last throughout one half of the wingbeat, i.e., for the whole upstroke or down- 
stroke, of a moth flapping its wings 30 times a second. The echo returning 
to the bat would be amplitude-modulated with a sharp peak at one point. At 
extreme range the brief peak would be the only part of the echo detected by the bat. 

Detection of an echo causes most bats to increase the repetition rate of their 
cries to as much as 150 per second. At the same time there is a decrease in the 
duration (to 1.0 millisecond or less) and in the frequency (to about 25 kcps) of 
each pulse of sound. Most noctuid moths have wingbeat frequencies of between 
10 and 40 per second. Therefore, when reception of an echo causes the bat to 
increase its pulse repetition rate a point must be reached where there is phasic 
interaction between pulse frequency and echo frequency determined by the moth's 
wingbeat. At some frequencies the phasing of pulse and echo source would 
produce a maximal echo every time, while at others the echo would be missed 
entirely. The signal significance of this effect, as well as the role played by 
frequency modulation in the bat's cry, cannot be estimated at present. 



The experimental work was supported by Grant E-947 from the United States 
Public Health Service to Tufts University. 

Some of the equipment was loaned by Professor Donald R. Griffin of Harvard 
University. The moths used were identified by Dr. Asher E. Treat of the 
City College of New York. 

SUMMARY 

1. Moths of several species were mounted in stationary flight and the 
echoes of ultrasonic pulses were recorded simultaneously with flash photographs 
of the attitude assumed by the wings at the instant the pulse reached the insect. 

2. The largest echo was produced by moths flying at the same altitude as 
the sound source when the potential course was roughly at right angles to the 
sound path and the wings were near to the top of the stroke. The difference 
between this maximum echo and that produced by the body and wings at other 
attitudes of the wing stroke was 30 decibels or more. 

3. Moths flying at the same altitude as the sound source but on other courses 
produced echoes that fluctuated with wing position. However, the maximum was 
never as great as that registered on the 90 course. 

4. Distortions in the shape of the echo are described and their causes are 
discussed. Scales on the wings or body of the moth do not appear to play an 
important anechoic role. 



210 KENNETH D. ROEDER 

5. It is concluded that the plane surface of the wings returns the major 
portion of the echo. The significance of this is discussed in relation to the 
problems of detection and evasion encountered under natural conditions by bats 
and flying moths. 

LITERATURE CITED 

GRIFFIN, DONALD R., 1958. Listening in the Dark. Yale University Press, New Haven, Conn. 
GRIFFIN, DONALD R., FREDERIC A. WEBSTER AND CHARLES R. MICHEL, 1960. The echolocation 

of flying insects by bats. Animal Behaviour, 8: 141-154. 
HASKELL, P. T., AND P. BELTON, 1956. Electrical responses in certain lepidopterous tympanic 

organs. Nature, 177: 139-140 
KOLB, A., 1958. Uber die Nahrungsaufnahme einheimischer Fledermause vom Boden. Verh. 

Dcutsch. Zool. Gcscllsch. iiu I-'rankjurt a.M., 162-168. 

MOHRES, F. P., 1950. Aus dem Leben unserer Fledermause. Kosmos, 46 : 291-295. 
ROEDER, KENNETH D., 1962. The behaviour of free flying moths in the presence of artificial 

ultrasonic pulses. Animal Behaviour, 10: 300-304. 
ROEDER, KENNETH D., AND ASHER E. TREAT, 1957. Ultrasonic reception by the tympanic 

organ of noctuid moths. /. Exp. Zool., 134: 127-158. 
ROEDER, KENNETH D., AND ASHER E. TREAT, 1961. The detection and evasion of bats by 

moths. Amer. Sci., 49: 135-148. (Reprinted in 1961 Annual Report of Smithsonian 

Institution.) 
TREAT, ASHER E., 1955. The response to sound in certain Lepidoptera. Ann. Ent. Soc. Amer., 

48: 272-284. 
WEBSTER, FREDERIC A., 1963. Bat-type signals and some implications. Human Factors in 

Technology (eds., E. Bennett, J. Degan, and J. Spiegel). McGraw-Hill Book Co., 

Inc., New York, N. Y. 
WEBSTER, FREDERIC A., AND DONALD R. GRIFFIN, 1962. The role of the flight membranes in 

the capture of insects by bats. Animal Behaviour, 10: 332-340. 




ENZYMIC HISTOCHEMISTRY OF GRANULAR COMPONENTS IN 
DIGESTIVE GLAND CELLS OF THE ROMAN SNAIL, 

HELIX POMATIA 1 

ROBERT M. ROSENBAUM AND BRUCE DITZION 

Department of Pathology, Albert Einstein College of Medicine, Nezv York 61, New York 

The digestive gland of Helix pomatia has been studied histologically especially 
during feeding and digestion (Krijgsman, 1925, 1929; Thiele, 1953; Guardabassi 
and Ferreri, 1953 among others). In addition, biochemical studies employing 
homogenates of gland tissue or aliquots of crop or digestive fluid have established 
that both intra- and extracellular hydrolytic enzymes are associated with the gland 
tissue (Holden and Tracey, 1950). Such methods, while providing quantitative 
data, do not permit direct study of the intracellular enzymic activity within the 
gland. Helix appears able to produce enormous amounts of several hydrolytic 
enzymes rapidly (Holden and Tracey, 1950; Billett, 1954; Dodgson and Powell, 
1959) and it is therefore not unreasonable to assume that a highly developed 
synthetic machinery for secretion of extracellular hydrolases may exist within the 
cells concerned. 

The activity of several hydrolytic enzymes, including acid phosphatase, /?-glu- 
curonidase, and several "cathepsins," has been shown by biochemical methods to 
vary according to the feeding cycle (Holden and Tracey, 1950; Jarrige and Henry, 
1952). Several histochemical studies have shown /?-glucuronidase activity (Billett 
and McGee-Russell, 1955) and acid and alkaline phosphatase activity (Guardabassi 
and Ferreri, 1953 ; Nakazima, 1956) within digestive gland tubules. To date, no 
reports have dealt with visualization of enzymic activity within specific digestive 
gland cells. 

In this investigation we were concerned not only with achieving intracellular 
localization of specific hydrolases, but also with a comparison of intracellular en- 
zymic activity during periods of starvation and active feeding. We were especially 
interested in the identification of enzymic activity with specific intracellular granules. 
The position of these granules has been shown to vary during feeding and digestion 
(Krijgsman, 1929; Rosen, 1941). With cytochemical methods, we were hopeful 
that it would be possible to extend these earlier cytological observations and estab- 
lish a more precise role for these granules consistent with some recent concepts of 
intracellular digestion and related hydrolytic enzymic activity. 

MATERIALS AND METHODS 

Our initial stock culture consisted of mature, estivating specimens of Helix 
pomatia from Morocco. 2 The snails were activated by exposure to a warm, humid 

1 This study was supported by grants from the United States Public Health Service 
<RG-S483 and A-360S) and a contract with the Office of Naval Research. 

2 Courtesy of Miscuraca Importing Corp., New York, N. Y. 

211 



212 ROBERT M. ROSENBAUM AND BRUCE DITZIOX 

environment. Animals were deemed "fed" after they had been observed to feed 
continuously on fresh lettuce leaves for several hours following starvation for 5-7 
days. Animals were considered "starved" when they had been isolated in indi- 
vidual fingerbowls for at least 7 days without food, or when they were killed while 
in estivation. For some experiments, animals were fed for at least 24 hours on 
fresh lettuce leaves soaked in a solution (25 mg. per 1 ml. water) of horse-radish 
peroxidase. 

The digestive gland was located by removing the apex of the shell and uncoiling 
the animal, the gland being identified as a green-brown mass near the upper end 
of the coil. For histochemical purposes, tissue was cut into small pieces (1-2 
mm.) immediately following removal and fixed in cold (4 C.) calcium-formalin 
(Baker's) or cold chloral hydrate formalin (Fishman and Baker, 1956) for 18-24 
hours. Other pieces of gland tissue were fixed in aqueous Benin's fluid, alcohol- 
formol-acetic acid fixative or Carney's fluid. 

Prior to being sectioned on a freezing microtome, digestive gland tissue must 
be embedded in gelatin, owing to its friable nature. Fixed tissue was washed briefly 
in cold water and placed for no longer than one hour in 15% gelatin at 37 C. 
and then hardened at C. for 20 minutes, and by additional treatment in cold 
10% neutralized formalin for one hour. Immediately prior to sectioning, the 
block was briefly rinsed in cold water. 

Enzymes: Acid phosphatasc activity was visualized by the lead-salt method 
of Gomori (1952 ), with /3-glycerophosphate as substrate, and by the azo dye method 
of Burstone ( 1958) with naphthol AS-MX phosphate as substrate and Red Violet 
LB as coupling reagent following cold acetone treatment (20 minutes) to remove 
lipid. With both methods, sections were incubated for 20-60 minutes at 37 C. 
(Rosenbaum and Rolon, 1962). For visualization of /3-ghtcnronidase activity, we 
employed cold chloral-hydrate-formalin-fixed tissue with 8-hydroxyquinoline glu- 
curonide as substrate (Fishman and Baker, 1956). Sections were incubated at 
37 C. for 30 minutes to 6 hours. For non-specific esterase activity, the para- 
rosanilin method of Lehrer and Ornstein (1959) was used following exposure of 
fixed sections to cold acetone. Incubation proceeded at C. for 30 minutes to 
two hours with a-naphthyl acetate as substrate. A parallel series of tissues was 
exposed to the organophosphorous compound E-600 (diethyl-/>-nitrophenyl phos- 
phate, 10~ n M in Tris-maleate buffer, pH 7.2) for one hour at 37 C. prior to 
incubation. This procedure has been considered capable of demonstrating type-C 
esterases (Hess and Pearse, 1958). For localization of aniinopcptidase activity, 
we employed the method of Burstone and Folk (1956), using the substrate L-leucyl 
/2-naphthylamide at pH 7.1 (0.2 M Tris buffer) following treatment of sections 
in cold acetone. Simultaneous coupling was obtained with the diazonium salt. 
Garnet GBC. 

Other histochemical methods: Detection of phospholipid was performed with 
the acid hematein method of Baker (1946). Some paraffin-embedded tissues fixed 
in Bouin's and Carnoy's fixatives were stained by the periodic acid Schiff method 
following digestion with salivary amylase. Finally, visualization of exogenous 
peroxidase, employed as a "marker" in some experiments, was accomplished by 
use of a hydrogen peroxide substrate and benzidine, essentially as described by 
Straus (1959). 



DIGESTIVE ENZYMES IN HELIX 



213 



RESULTS 

Cytology of the digestive gland: Several authors have presented good descrip- 
tions of the kinds of cells in the digestive gland of Helix (Krijgsman, 1929; Rosen, 
1941; Thiele, 1953). This work has resulted in a variety of terms frequently 
descriptive of the same cell type. It is desirable, therefore, to clarify our nomen- 
clature for the cell types considered in the present study. 



S.R. 



B. 



cc.gr 



yel.gr 




c.c. gr. 
yel.gr. 




hem. space 




feeding 

FIGURE 1. A. Schematic drawing of a digestive gland tubule from Hcli.r pomatia, show- 
ing relationships of the different cytological components. B. Distribution of various intracellu- 
lar granules during the feeding stages of Helix. Clear colorless granules are almost exclusively 
localized at the luminal border of the secretory-resorption cells. C. In secretory-resorption 
cells from starved animals, the clear, colorless granules are not concentrated at the luminal 
border. CA, calcareous cell ; SR, secretory-resorption cell ; c.c., gr., clear, colorless granules ; 
yel. gr., yellow granules ; b.m., basement membrane ; hem. space, hemocoelic space. 



Our observations concern two kinds of cells (Fig. 1) the so-called calcium 
cell and the digestive or SR cell (secretory-resorption cell Rosen, 1941). The 
calcium cell is characterized by its triangular shape with the broad base (approxi- 
mately 50 /A) touching the basement membrane. The cell points toward, but 
does not extend into, the tubular lumen. The cytoplasm contains numbers of large, 
nearly colorless spherules, the so-called calcium spherules. Secretory-resorption 
(SR) cells are greater in number than, and adjacent to, the calcium cells. The 
SR cells are long (100 /*), the body of the cell extending from the basement mem- 



214 



ROBERT M. ROSENBAUM AND BRUCE DITZION 



brane into the lumen of the glandular tubule. The nucleus is generally smaller 
than that of the calcium cell. The SR cell possesses granular inclusions, the num- 
ber and distribution of which vary considerably with the feeding stage of the 
animal (Krijgsman, 1929). These granules are generally described according to 
three types: large brown granules not present in all cells, but, when present, con- 




FIGURE 2. Three calcium cells (arrows) from the digestive gland of a starved snail, 
stained with the method of Gomori, omitting glyercophosphate from the incubation medium. 
Calcium spherules are stained by lead sulfide. All other regions of the digestive gland are 
negative. 360 X. 

FIGURE 3. Four calcium cells from the digestive gland of a recently fed snail, stained 
as in Figure 2 but with glycerophosphate present as substrate. Calcium spherules, stained by 
the false positive reaction, are obscured by an intense cytoplasmic staining due to enzymic 
activity. A few spherules have been scattered extracellular ly in processing. 360 X. 

FIGURE 4. Tubule from the digestive gland of a starved animal, stained for acid phospha- 
tase activity by the azo dye method of Burstone. Enzymic activity is present in calcium cells 
(arrows), and weaker activity is present in the cytoplasm of SR cells. Spherules do not 
stain. Green filter. 360 X. 

FIGURE 5. Tubule from the digestive gland of a recently fed animal, stained as in Figure 
4. Intense enzyme activity is seen in the cytoplasm of three calcium cells. The spherules do 
not stain. The cytoplasm of the SR cells is heavily stained, especially near the lumen 
(arrows). Green filter, 360 X. 



DIGESTIVE ENZYMES IN HELIX 



215 



tained within vacuoles located at the base of the cell in the vicinity of the nucleus; 
small yellow granules generally distributed throughout the cytoplasm ; small clear 
granules apparently free in the cytoplasm near the periphery of the cell. The dis- 
tribution of all granule types is summarized in Figure 1A. 

Enzyme distribution within the gland cells of starved and feeding animals 

Acid phosphatase: With the Gomori metal-salt method, we could not distinguish 
acid phosphatase activity in the cytoplasm of calcareous cells since calcium spherules 
from both starved and feeding animals blackened intensely with the sulfide. These 
calcium granules were large, round bodies, which frequently became dispersed 





v-: 



FIGURE 6. Digestive gland tubules from a starved animal, stained for J-glucuronidase 
activity by the method of Fishman and Baker, following fixation in cold chloral-hydrate- 
formalin. Enzymic activity is confined to the yellow granules. Brown granules (arrows) 
did not stain for enzymic activity but are emphasized by the yellow filter employed. Calcium 
cells are negative. 200 X. 

FIGURE 7. Digestive gland tubule from a recently fed animal, stained as in Figure 6. 
There is increased cytoplasmic activity, especially at the periphery of SR cells. The yellow 
granules are also stained. Calcium cells are negative. Yellow filter. 200 X. 

outside the cell due to compression of the sections during processing (Figs. 2 and 
3). Application of heat to the sections (90 C. for at least 10 minutes), or omis- 
sion of glycerophosphate substrate from the incubation medium, demonstrated that 
this localization of final reaction product was a false positive reaction not due to 
enzymic activity. With use of an azo dye method for acid phosphatase activity, 
the spherules showed no enzymic activity (Figs. 4 and 5). With staining by the 
azo dye method, however, it was clear that the cytoplasm of calcareous cells from 
recently fed animals showed increased acid phosphatase activity over cells from 
starved animals (compare Figs. 4 and 5). 

The SR cells from starved animals showed little or no enzymic activity by 
either the azo dye or lead-salt methods. In fed animals, some enzymic activity 
could be detected near the luminal border in the region where the small, colorless 



216 



ROBERT M. ROSENBAUM AND BRUCE DITZIOX 




FIGURE 8. Digestive gland tubule from a starved animal, stained for non-specific esterase 
activity by the method of Lehrer and Ornstein. There is intense cytoplasmic staining, and 
staining of granules is evident in SR cells. Calcium cells (arrows) are negative. Green 
filter, 380 X. 



DIGESTIVE ENZYMES IN HELIX 217 

granules accumulated. Neither the small yellow nor large brown granules pos- 
sessed demonstrable acid phosphatase activity by the methods we employed. 

When sections were exposed to the non-ionic surface-activating agent, Triton 
X-100 (0.25% at 4 C. for one hour ), a distinct decrease of enzymic activity within 
calcium cells could be detected with the azo dye method, especially in tubules from 
starved animals. With identical treatment applied to tissue from feeding animals, 
enzymic activity in calcium cells and at the lumen of SR cells was also diminished. 

fl-Glucuronidase (p-c/lucosidiironidasc ) : Calcium cells showed no activity for 
this enzyme. In SR cells from starved animals, intracellular enzymic activity could 
be detected in locations approximating that of the yellow granules (Fig. 6). The 
large brown granules showed no enzymic activity. Diffuse staining of the peripheral 
cytoplasm occurred in locations consistent with those of the small colorless granules. 
In secretory-resorption cells from fed animals, staining for /?-glucuronidase activity 
was intense and generally distributed throughout the cytoplasm. Activity for this 
enzyme also appeared in the yellow granules (Fig. 7). 

Esterases: Calcium cells from both starved and feeding animals showed no 
esterase activity (Figs. 8 and 9). Secretory-resorption cells stained intensely for 
non-specific esterase activity. Activity appeared throughout the cytoplasm (Fig. 8), 
but with fed animals the reaction \vas more intense and diffuse (Fig. 10) than 
with starved animals. In all cases, staining for non-specific esterase activity could 
not be identified with specific granules. 

In both starved and feeding animals, treatment with the organo-phosphorous 
inhibitor E-600 resulted in a considerable loss of cytoplasmic activity. In prepara- 
tions from starved animals, E-600-resistant esterase activity was limited to what 
appeared to be the yellow granules of the SR cells (Fig. 9). Treatment of sections 
from fed animals in E-600 resulted in a diffuse localization of enzymic activity 
within peripheral regions of SR cells (Fig. 11). This region generally contained 
the small, colorless granules. 

Aininopcptidasc : Calcium cells from both starved and feeding animals showed 
no enzymic activity. Some slight activity was detected in secretory-resorption cells 
from starved and feeding animals. 

Other histocheuiical inctJwds 



With Baker's acid hematein method for phospholipid, positive 
staining occurred in the peripheral region of SR cells from starved and feeding 
animals (Figs. 12 and 13). Localization of phospholipid in SR cells in the glands 
of starved animals was more diffuse (Fig. 12), while in fed animals it was limited 

FIGURE 9. Digestive gland tubule from a starved animal, stained as in Figure 8, following 
treatment of the section in the organo-phosphorous inhibitor E-600. Non-specific enzyme 
activity present in the cytoplasm of cells in Figure 8 has been removed but enzymic activity 
remains in yellow granules scattered throughout the SR cells. Calcium cells are negative. 
Green filter. 380 X. 

FIGURE 10. Digestive gland tubule from a feeding animal, stained for non-specific esterase 
activity as in Figure 8. There is intense enzyme activity in SR cells. Green filter. 380 X. 

FIGURE 11. Digestive gland tubule from a recently fed animal. The section was treated 
with E-600 and stained for esterase activity as in Figure 9. While general cytoplasmic 
enzyme activity is diminished, there is positive staining for E-600-resistant enzyme activity 
near the lumen of the tubule (arrows). The calcium cells are negative. Green filter. 380 X. 



218 



ROBERT M. ROSENBAUM AND BRUCE DITZION 

> 




FIGURE 12. Tubule from the digestive gland of a starved animal, stained for phospholipid 
by the method of Baker. There is peripheral staining of SR cells, both at the luminal border 
(arrows) and the lateral cell margins. Brown granules, although detectable, are not stained 
with this method. 360 X. 



DIGESTIVE ENZYMES IN HELIX 219 

to the luminal border (Fig. 13). The location of phospholipid staining corre- 
sponded to areas containing the small, colorless granules. A weak reaction for 
phospholipid was present throughout the cytoplasm of SR cells in both starved 
and feeding animals. There was no staining of the calcium cells. 

PAS reaction: The periodic acid Schiff reaction was employed following incu- 
bation with amylase at pH 7.4 for digestion of glycogen. In secretory-resorption 
cells from starved animals, the yellow granules were intensely Schiff-positive (Fig. 
14). In cells from fed animals, fewer yellow granules were present and these 
showed a diminished Schiff reaction (Fig. 15). In glands from fed animals fixed 





FIGURE 16. Digestive gland tubule from a snail fed on lettuce soaked in horse-radish 
peroxidase, visualized by means of a benzidine reaction. Numerous "vacuoles" with enzyme 
activity appear throughout SR cells. Brown granules are also stained by the benzidine reaction 
but do not possess enzyme activity. Yellow filter. 360 X. 

in aqueous Bouin's fluid, a weak Schiff-positive reaction appeared at the luminal 
border of the secretory-resorption cells (Fig. 15). This region corresponded to 
where localized phospholipid was detected and where, in fed animals, the small 
colorless granules were always observed. 

Ingestion of peroxidase: Within 15 minutes following initial attachment of a 
snail onto a leaf of treated lettuce, peroxidase could be detected in the cytoplasm 

FIGURE 13. Tubule from the digestive gland of a fed animal stained for phospholipid as 
in Figure 12. The reaction is present chiefly at the luminal border of SR cells. 360 X. 

FIGURE 14. Tubule from the digestive gland of a starved snail, fixed in Bouin's fluid and 
stained with periodic acid Schiff method following digestion with saliva. The yellow granules 
within SR cells appear intensely Schiff-positive. Green filter. 360 X. 

FIGURE 15. Digestive gland tubule from a fed animal, stained as in Figure 14. A few 
yellow granules are present but weakly stained. There is staining of the luminal border of 
SR cells, corresponding to the location of clear, colorless granules (arrows). Green filter, 360 X. 



!20 ROBERT M. ROSENBAUM AND BRUCE DITZION 

of the secretory-resorption cells (Fig. 16). Localization of ingested enzyme ap- 
peared to be within ''vacuoles" scattered throughout the cytoplasm, especially 
near the luminal border of the SR cells. The location of the larger peroxidase- 
positive "vacuoles" corresponded to that of the yellow granules. With extended 
feedings (up to three hours) on treated lettuce, we could detect no increase in 
the amount of peroxidase ingested, nor did existing sites of peroxidase activity 
coalesce to form larger vacuoles. 

Control experiments on cells from animals fed untreated lettuce revealed no 
peroxidase activity endogenous to the snail itself. 

DISCUSSION 

In this study, hydrolase activity was localized within secretory-resorption cells 
and calcium cells from the digestive gland of Helix. The location of some of this 
enzymic activity closely corresponded to the distribution of migrating intracellular 
granules described by Krijgsman (1925, 1929) and by Rosen (1941). 

The activities of the hydrolases considered here have previously been visualized 
primarily in cells from vertebrate tissues, where their role in resorption phenomena 
or intracellular digestion has been described. Thus, acid phosphatase activity has 
been identified within phagocytic macrophages (Weiss and Fawcett, 1953 ), with iron 
resorption in liver (Novikoff and Essner, 1960) and protein resorption in kidney 
(Straus, 1961). The activity of two other hydrolases has been visualized in the 
peribiliary region of vertebrate liver cells (esterase Holt, 1956; /3-glucuronidase 
Goldfarb and Barka, 1960), where their role in resorption was suggested. Studies 
with invertebrates have identified intracellular hydrolytic activity with feeding and 
digestion in planarians (Rosenbaum and Rolon, 1960), amebas (Novikoff, 1959; 
Birns, 1960) and ciliate protozoa (Seaman, 1961; Rosenbaum and Wittner, 1962). 

Unfortunately, those few studies describing the activity of hydrolytic enzymes 
in molluscan digestive glands (see review of Arvy, 1962) employed methods no 
longer deemed completely reliable for purposes of intracellular localization. In 
the present investigation, we attempted to increase reliability of the methods by 
minimizing enzyme diffusion and inactivation through use of frozen sections of tissue 
fixed briefly in the cold. Such methods, especially in combination with newer 
naphthol substrates and rapid simultaneous coupling to a diazonium salt, offered 
additional protection against diffusion of enzyme, final colored reaction product, 
or both (Rosenbaum, 1962). Control of these factors served to support our obser- 
vations that much, if not all, of the intracellular enzymic activity we studied was 
issociated with granular or with vacuolar structures within the cells concerned. 

Within a short time following ingestion of food, the secretory-resorption cells 
showed increased activity for several enzymes, especially /8-glucuronidase and es- 
terases. Much of this activity was associated with yellow granules, although non- 
-pecific esterase activity was also diffusely distributed in the cytoplasm of secretory- 
resorption cells. It is noteworthy that the yellow granules also served as sites of 
exogenous horse-radish peroxidase accumulation. We propose that this enzyme 
entered the cell by pinocytosis, perhaps via the clear, colorless granules, which 
may represent small pinocytic vacuoles. The observation that these small granules 
stained for phospholipid and were also Schiff-positive suggests a possible relation- 



DIGESTIVE ENZYMES IN HELIX 221 

ship to the cell membrane. During feeding, the region of accumulation of these 
clear, colorless granules also possessed increased E-600-resistant esterase activity. 

The present observations did not permit determination as to whether the yellow 
granules in the secretory-resorption cells could form vacuolar structures associated 
with the granules as resorption took place. The accumulation of pinocytic vacuoles 
associated with migrating granular structures has been described in specialized 
vertebrate cells grown in tissue culture (Rose, 1957a, 1957b). In Paramecium, 
Rosenbaum and Wittner (1962) have described migration of neutral red-staining 
granules from a region beneath the pellicle toward forming food vacuoles as these 
become larger and begin to show increased activity for acid phosphatase, non- 
specific and E-600-resistant esterases during feeding. In addition to this enzymic 
activity, the neutral red bodies of the paramecium stained for phospholipid, and 
their possible relationship to vertebrate hepatic lysosomes (deDuve, 1959) was 
therefore suggested. We have made no observations with respect to the ability of 
granular components within the SR cells of Helix to stain selectively with neutral 
red. 

The present study points also to the possibility that intracellular digestion may 
take place within secretory-resorption cells of the digestive gland. Phagocytosis, 
which is closely related to pinocytosis, occurs in Helix (Krijgsman, 1929; Rosen, 
1941; van Weel, 1961), although its role in feeding or digestion has been little 
studied. However, a pinocytic mechanism, permitting active absorption of dis- 
solved food substances, would be a more attractive one than simple "diffusion," 
the process suggested by earlier investigators (Hirsch, 1915; Jordan and Bege- 
mann, 1921). There is no question that the first stage of digestion in the snail 
depends on a large number of enzymes acting in the gut lumen (Holden and 
Tracey, 1950; van Weel, 1961). Such extracellular enzymic activity could serve 
to break down ingested food initially, thereby permitting further intracellular diges- 
tion to take place. 

Our present study does not test the possibility that migration of the clear color- 
less granules to the luminal border of the secretory-resorption cells during feeding 
could be related to discharge of enzymes into the lumen. The presence of increased 
E-600-resistant esterase activity at the luminal border during feeding might repre- 
sent a stage in the release of proteolytic enzymes into the glandular lumen. It 
would seem, however, that initially extracellular digestion in the gut of an herb- 
ivorous species such as Helix would be brought about primarily by carbohydrases, 
not by enzymes more specifically suited to protein hydrolysis. It has been suggested 
(van Weel, 1961) that the salivary glands chiefly produce carbohydrases, and 
even that some of these enzymes are not produced by the animal itself (Florkin 
and Lozet, 1949; Jeuniaux, 1954). Once food has reached the gut and digestion 
begun, however, succeeding steps must involve some degree of intracellular diges- 
tion, especially of protein. Activity of the intracellular hydrolases described in the 
present study would appear to be well suited for intracellular breakdown of plant 
protein. 

SUMMARY AND CONCLUSIONS 

1. Cytochemical visualization methods for activity of acid phosphatase, /?-glu- 
curonidase, aminopeptidase and non-specific and E-600-resistant esterases were 



ROBERT M. ROSENBAUM AND BRUCE DITZION 

applied to digestive gland tissue from starved and feeding Helix pomatia. Other 
cytochemical methods used included Baker's acid hematein for phospholipid and 
the periodic acid Schiff method. 

2. Calcareous cells stained only for acid phosphatase activity, by both lead-salt 
and an azo dye method. Calcium granules within these cells did not stain with 
the azo dye method while false-positive reactions in the granules were always 
obtained with the lead-salt method. Some increase in enzymic activity was detected 
in the cytoplasm of feeding animals. Secretory-resorption (SR) cells showed little 
activity for acid phosphatase. 

3. The yellow granules in SR cells stained for /3-glucuronidase activity in both 
starved and feeding animals. After feeding, SR cells showed an increase in 
enzymic activity, both in granules and more diffusely in the cytoplasm. Although 
some aminopeptidase activity was present, insignificant differences in location and 
intensity of the enzyme were detected. 

4. SR cells stained intensely for non-specific esterase activity. Cytoplasmic 
staining for this class of enzymes was intense in feeding animals. Treatment with 
the inhibitor E-600 resulted in loss of cytoplasmic staining; activity persisted in 
yellow granules in both starved and feeding animals. Some activity was detected 
in small, colorless granules. 

5. Phospholipid was detected in peripheral regions of the SR cells from starved 
and feeding animals. The reaction was diffuse in starved animals but concentrated 
at the luminal border of the SR cells in fed animals. Yellow granules of SR cells 
were periodic acid Schiff-positive. Especially in fed animals, periodic acid Schiff 
granules appeared at the luminal border of SR cells. Starved animals fed on let- 
tuce leaves impregnated with horse-radish peroxidase showed accumulation of 
enzyme in vacuoles closely associated with the yellow granules. 

6. The observations extend the concepts, advanced by earlier workers, that 
granular components of the secretory-resorption cells play a significant role in 
digestion in the snail. The location of activity of the several hydrolases studied, 
and the alterations in response to feeding, suggest that these classes of enzymes and 
the granules form a functioning unit in the physiology of intracellular digestion 
in Helix. 

LITERATURE CITED 

ARVY, L., 1962. Histochemie des enzymes impliques dans la digestion, dans la serie animale. 

Handbk. Histochem. VII, 2: 154-303. 

BAKER, J. R., 1946. Histochemical recognition of lipine. Quart. J. Micr. Sci., 87 : 441-470. 
BILLET, F., 1954. The /3-glucuronidase of the Roman snail (Helix pomatia). Biochem. J., 

57: 159-162. 

BILLET, F., AND S. M. McGEE-RussELL, 1955. The histochemical localization of /3-glucu- 
ronidase in the digestive gland of the Roman snail (Helix pomatia). Quart. J. Micr. 

Sci., 96: 35-48. 
BIRNS, M., 1960. The localization of acid phosphatase activity in the amoeba, Chaos chaos. 

Exp. Cell Res., 20 : 202-205. 
BURSTONE, M. S., 1958. Histochemical demonstration of acid phosphatase with naphthol-AS 

phosphates. /. Nat. Cancer Inst., 21 : 523-540. 
BURSTONE, M. S., AND J. E. FOLK, 1956. Histochemical identification of aminopeptidase. 

/. Histochem. Cytochem., 4: 217-226. 
DODGSON, K. S., AND G. M. POWELL, 1959. Studies on sulphatases. 26. Arylsulphatase 

activity in the digestive juice and digestive gland of Helix pomatia. Biochem. J., 

73: 666-671. 



DIGESTIVE ENZYMES IN HELIX 223 

DEDuvE, C, 1959. Lysosomes, a new group of subcellular particles. In "Subcellular Particles" 

(ed. by T. Hayashi), pp. 128-158. Ronald Press, New York. 
FISHMAN, W. H., AND J. R. BAKER, II, 1956. Cellular localization of /3-glucuronidase in rat 

tissue. /. Histochem. Cytochem., 4 : 570-587. 
FLORKIN, M., AND F. LOZET, 1949. Origine bacterienne de la cellulase du contenu intestinal 

de 1'escargot. Arch. int. Physiol., 57 : 201-207. 
GOLDFARB, S., AND T. BARKA, 1960. Cytological localization of /3-glucuronidase. /. Histochem. 

Cytochem., 8: 226-227. 

GOMORI, G., 1952. Microscopic Histochemistry. Chicago : University of Chicago Press. 
GUARDABASSI, A., AND E. FERRERI, 1953. Istofisiologia dell'apparato digerente di Helix 

pomatia. Arch. Zool. Ital., 38: 63-156. 
HESS, R., AND A. G. E. PEARSE, 1958. The histochemistry of indoxylesterase of rat kidney 

with special reference to its cathepsin-like activity. Brit. J. Exp. Path., 39 : 292-299. 
HIRSCH, G. C., 1915. Ernahrungsbiologie fleischfressender Gastropoden. Zool. Jahrb. Physiol., 

35 : 357-504. 
HOLDEN, M., AND M. V. TRACEY, 1950. A study of enzymes that can break down tobacco-leaf 

components. 2. Digestive juice of Helix on defined substrates. Biochem. J., 47 : 

407-414. 

HOLT, S. J., 1956. The value of fundamental studies of staining reactions on enzyme histo- 
chemistry with reference to indoxyl methods for esterases. /. Histochem. Cytochem., 

4: 541-554. 
JARRIGE, P., AND R. HENRY, 1952. fitude de 1'activite glucuronidasique du sue digestif d'Helix 

pomatia L. et de sa phenolsulfatase. Bull. Soc. Chim. Biol., 34 : 872-885. 
JEUNIAUX, C., 1954. Sur la chitinase et la flore bacterienne intestinale des mollusques 

gasteropodes. Bull. Acad. Roy Belg., CI, Sci. Mem. 28. F7. 
JORDAN, H. J., AND H. BEGEMANN, 1921. Ueber die Bedeutung des Darnies von Helix 

pomatia. Zool. Jahrb. Physiol., 38 : 565-582. 
KRIJGSMAN, B. J., 1925. Arbeitsrhythmus der Verdauungsdriisen bei Helix pomatia. I. 

Zeitschr. vergl. Physiol, 2: 264-296. 
KRIJGSMAN, B. J., 1929. Arbeitsrhythmus der Verdauungsdrusen bei Helix pomatia. II. 

Zeitschr. vergl. Physiol., 8: 187-280. 

LEHRER, G. M., AND L. ORNSTEIN, 1959. A diazo coupling method for the electron micro- 
scopic localization of cholinesterase. /. Biophys. Biochem. Cytol., 6 : 399-406. 
NAKAZIMA, M., 1956. On the structure and function of the midgut gland of mollusca, with a 

general consideration of the feeding habits and systematic relations. Jap. J. Zool., 

11 : 469-566. 
NOVIKOFF, A. B., 1959. Lysosomes and the physiology and pathology of cells. Biol. Bull., 

117: 385. 
NOVIKOFF, A. B., AND E. ESSNER, 1960. The liver cell : some new approaches to its study. 

Amer. J. Med., 29 : 102-131. 
ROSE, G. G., 1957a. Microkinetospheres in VP satellites of pinocytic cells observed in tissue 

culture of Gey's strain HeLa with phase contrast cinematographic techniques. /. 

Biophys. Biochem. Cytol., 3: 697-704. 
ROSE, G. G., 1957b. Observations on the dynamics of pinocytic and varient pinocytic (VP) 

cells in Gey's human malignant epidermoid strain HeLa. Texas Repts. Biol. Mcd., 

15: 313-331. 
ROSEN, B., 1941. Intrazellulare Verdauung bei Helix pomatia L. Zool. Jahrb. Abt. f. allg. 

Zool. u. Physiol., 60: 241-252. 
ROSENBAUM, R. M., 1962. The intracellular localization of enzymes. Handbk. Histochem. 

VII, 4: 1-108. 
ROSENBAUM, R. M., AND C. I. ROLON, 1960. Intracellular digestion and hydrolytic enzymes 

in the phagocytes of planarians. Biol. Bull., 118: 315-323. 
ROSENBAUM, R. M., AND C. I. ROLON, 1962. Species variability and the substrate specificity 

of intracellular acid phosphatases : a comparison of the lead-salt and azo-dye methods. 

Histochem., 3 : 1-16. 

ROSENBAUM, R. M., AND M. WITTNER, 1962. Intracytoplasmic particles and enzymes associ- 
ated with feeding and digestion in Paramccium caudatum. The possible role of the 

neutral red granules. Arch. f. Protistenk., 106: 223-240. 



224 ROBERT M. ROSENBAUM AND BRUCE DITZION 

SEAMAN, G. R., 1961. Acid phosphatase activity associated with phagotrophy in the ciliate 

Tetrahymena. J. Biophys. Biochem. Cytol., 9: 243-245. 
STRAUS, W., 1959. Regional cytochemical identification of phagosomes in various tissues of 

the rat and their differentiation from mitochondria by the peroxidase method. /. 

Biophys. Biochem. Cytol., 5: 192-204. 
STRAUS, W., 1961. Cytochemical observations on the transport of intravenously injected 

horse-radish peroxidase and the development of phagosomes in the cells of the kidney 

of the rat. Ex p. Cell Res., 22 : 282-291. 
THIELE, G., 1953. Vergleichende Untersuchungen iiber den Feinbau und die Funktion der 

Mitteldarmdriise einheimischer Gastropoden. Zcitschr. Zcllfnrscli., 38: 87-138. 
VAN WEEL, P. B., 1961. The comparative physiology of digestion in molluscs. Amer. Zool., 

1: 245-252. 
WEISS, L. P., AND D. W. FAWCETT, 1953. Cytochemical observations on chicken monocytes, 

macrophages and giant cells in tissue culture. /. Histochem. Cytochem., 1 : 47-65. 



OSMOTIC REGULATION IN MARINE AND FRESH-WATER 
GAMMARIDS (AMPHIPODA) 1 - 2 

HENRY O. WERNTZ3 

Department of Biology, Northeastern University, Boston 15, Massachusetts 

Different species of the amphipod genus Gammarus live in habitats with a 
wide range of salt concentrations, from fresh water to sea water. To withstand 
such diverse conditions, the species must differ at least quantitatively in their 
osmotic physiology. In comparing four British species of the genus, Beadle and 
Cragg (1940a) found that the salinity range which each species tolerated cor- 
related with the level at which it regulated the chloride concentration and the total 
osmotic concentration of its blood. More recently, Shaw and Sutcliffe (1961) 
and Lockwood (1961) have analyzed some mechanisms responsible for the osmo- 
regulatory differences between two of Beadle and Cragg's species. 

The present paper, too, attempts to analyze osmoregulatory mechanisms of 
gammarids, using an approach rather different from those of the previously men- 
tioned investigators. The work involves four previously unstudied species : three 
in the genus Gammarus and one in the closely related genus Marinogainumrus. 
The species are first compared with respect to the blood concentrations which they 
maintain in various dilutions of sea water. Then the role of the nephridium in 
salt and water excretion is studied in two species : one marine and one fresh-water. 
The results obtained imply some specific differences in the ionic uptake mechanisms. 

EXPERIMENTAL MATERIAL 
Adult males of the following four species were used: 

Gammarus oceanicus Segerstrale (1947), characterized in the original 
description as "mainly marine," is widely distributed along both shores of the 
more northern parts of the Atlantic Ocean. It also lives in brackish situations 
and has been found in the Gulf of Finland in salinities as low as 2.5%o. Originally 
described as a subspecies of G. saddachi Sexton, it was elevated to specific rank 
by Kinne (1954). The animals used in my experiments (mostly 18 to 23 mm. 
in length and 110 to 220 mg. in weight) were collected from an intertidal fresh- 
water seep on Cape Cod Bay, about two miles north of the Cape Cod Canal. 

1 Presented to the Faculty of Yale University in partial fulfillment of the requirements 
for the degree of Doctor of Philosophy. 

- Contribution number 1291, Woods Hole Oceanographic Institution. 

3 Part of the work reported here was undertaken while I was recipient of a National 
Science Foundation Graduate Fellowship, and part while I was research assistant at the 
Woods Hole Oceanographic Institution. I wish to express special indebtedness to my thesis 
director, Professor G. E. Hutchinson, whose suggestions led to the present researches, and 
to Mr. H. J. Turner, who made available the facilities of his laboratory at the Woods Hole 
Oceanographic Institution and who aided the work in innumerable ways. 

225 



226 HENRY O. WERNTZ 

Marinogammarus finmarchicus (Dahl) was characterized by Sexton and 
Spooner (1940, p. 659) as "a marine littoral species, occurring intertidally along 
the sea coast and for a short distance up estuaries." The genus was formerly 
( Schellenberg, 1937) a subgenus of Gammarus. The animals (about 15 to 20 
mm. long) were collected from a rocky shore at Manomet, Massachusetts, on 
Cape Cod Bay. 

Gammarus tigrinus Sexton (1939) is a brackish-water species very similar in 
appearance to G. fasciatus. For a time the two names were considered synonymous. 
That the forms are actually separate species is indicated by the following evidence : 
there are some slight but consistent morphological differences (Bousfield, 1958) ; 
the osmotic physiology of the two forms differs, as shown later in this paper; 
and in the culture experiments I performed, the two bred separately but did not 
cross-breed. The experimental animals came from a fresh-water seep high in 
the intertidal zone on Nonamesset Island, in Vineyard Sound, Massachusetts. 
They were 10 to 13 mm. long. 

Gammarus jasciatiis Say is the common fresh-water species in New England 
and was taken from various streams near Woods Hole, Massachusetts, and New 
Haven, Connecticut. The animals were 10 to 15 mm. in length and 25 to 60 mg. 
in weight. 

METHODS 

Determining osmotic concentration 

The osmotic concentration of fluids was determined from the freezing point 
depression. Actually, the melting point of solutions was determined by observing 
a frozen sample as it was slowly warmed. 

The apparatus used is a modification of the one designed by Kinne (1952) 
and is also similar to one devised by Ramsay (1949). It consisted of an aquarium 
of copper and glass which was mounted in the front wall of an insulated box and 
filled with a 25% solution of ethyl alcohol in water. The continuously stirred 
bath was warmed by an immersion heater, which was regulated by a variable 
transformer. The bath was cooled by a pack of crushed dry ice, which was pressed 
against the back of the aquarium by a spring-driven plunger. By careful balancing 
of the heat source and sink the temperature could be accurately controlled. The 
temperature was read with a Heidenhain thermometer graduated to 0.01 C. 

To obtain a blood sample, the animal was gently dried off between layers of 
absorbent paper toweling, and its back was wiped with filter paper moistened with 
distilled water. With the animal held under paraffin oil the exoskeleton over the 
heart was punctured with a fine needle. From the drop of blood which appeared, 
a sample was taken with a fine Pyrex capillary tube, which had been cleaned in 
hot nitric acid. First, paraffin oil was drawn up into the capillary, then a small 
sample, and finally more paraffin oil. About a centimeter of capillary containing 
the sample was broken off, the ends sealed with sealing wax, and a label affixed. 
The samples were stored in a freezer. The size of the sample drops was not pre- 
cisely controlled, but usually was 0.1 to 0.2 mm. in diameter and 0.3 to 1 mm. in 
length. This gave a range in volume from about 0.003 to 0.03 mm 3 . 

To determine the freezing point the sample was first frozen in crushed dry ice 
and then transferred to the controlled temperature bath, which had been cooled 



OSMOTIC REGULATION IN GAMMARIDS 227 

below the expected freezing point of the sample. As many as six samples were 
held simultaneously in the bath by a movable clamp. The samples were illuminated 
by light passed through a square of Polaroid film and reflected from a mirror 
behind them. They were watched through the front glass wall of the aquarium 
with a horizontally mounted microscope containing a Polaroid analyzer. Since 
ice crystals are birefringent, use of polarized light made the crystals easier to see 
as brilliant white objects against a dark background. 

The temperature of the bath was raised fairly rapidly until most of the ice 
in the sample had melted. The rate of temperature rise was then slowed to 0.01 C. 
per minute, or less. At this slow rate it could be assumed that the ice and the 
melted solution were nearly in thermodynamic equilibrium, and that the last crystal 
disappeared at the freezing point, provided that the rate of temperature rise had 
been slowed five minutes or more before this event. The method was accurate 
within 0.01 C. on standard salt solutions. In duplicate blood samples from a series 
of animals the maximum difference in the freezing point was 0.02 C. 

Since a major concern was with the osmotic movement of water, it was deemed 
most suitable to express the freezing point depressions and the salinities of the 
media as moles of ideal non-electrolytic solute per kilogram of water. The equiva- 
lent in moles of monovalent salt, such as NaCl, may be found by dividing by 2. 
Use of molal units has the additional convenience that the osmotic concentration of 
"normal" sea water (chlorinity of 19.4'/c ) is nearly unity: 1.03 molal. Hence the 
molality of a solution is nearly equal numerically to its fraction of sea-water strength. 

Determining osmoregulatory beJiavior of species 

Animals were individually isolated without food in about 200 ml. of medium 
in one-pint polyethylene boxes during the period of adaptation to a new salinity. 
It was experimentally determined that the blood concentration of Gammarus oceani- 
cus reached a new steady-state about 12 hours after transfer from undiluted sea 
water (0.93 molal at Woods Hole) to 0.1 molal sea water. G. fasciatus reached 
a new steady-state within 1^ hours after transfer from fresh water to 0.6 molal 
sea water, the highest concentration in which it normally survived. Although 
adaptation times to other salinities were not determined, these experiments are 
taken to indicate that the one- or two-day period of adaptation was appropriate. 

Animals were transferred to fresh water and to sea water of molality 0.03, 
0.1, 0.2, 0.4, 0.61, 0.82, 0.93 or 1.03, and, in some instances, 1.5. (These concentra- 
tions correspond to salinities of 1.0, 3.5, 7.0, 14, 21, 28, 32 or 35, and 51.5%o.) 
All the species were exposed to the experimental media for 48 hours, or a little 
longer, with the exception of G. oceanicus, which was exposed for only 24 hours. 
Owing to the method of sampling, each animal could be used only once ; hence 
the data shown on the curves are composite. The total number of surviving animals, 
distributed more or less equally among the various media, was 51 for G. oceanicus, 
74 for M. finmarchicus, 77 for G. tigrinus, and 64 for G. fasciatus. The tempera- 
ture range during the experiments was 16 to 19 C. for G. oceanicus, 14 to 16 C. 
for M. finmarchicus and G. tigrinus, and 13 to 18 C. for G. fasciatus. The curves 
were determined in July, 1955, for G. oceanicus, in September, 1955, for G. fasciatus, 
and in July, 1956, for G. tigrinus and M. finmarchicus. 

No attempt was made to determine or control the molting stage of the experi- 



HENRY O. WERNTZ 

mental animals. Baumberger and Olmsted (1928) found that the concentration of 
the blood in some brackish-water crabs doubled during a molt. In my experiments 
the blood concentration in any one medium varied only slightly. Either a concen- 
tration change does not occur during molt in these animals or it is transitory and 
was not encountered. 

Determining urinary rate and concentration 

The structurally simple nephridium of Gammarus has an end-sac and a more 
or less coiled canal but no storage bladder (Burian and Muth, 1924; Schwabe, 
1933). This arrangement should result in a continuous flow of urine, rather than 
intermittent micturition. The nephridium opens at the tip of a protuberance (the 
nephrocone) on the second segment of the second antenna; consequently the open- 
ing is readily accessible. 

Sampling the urine was first attempted by immersing the animal in paraffin 
oil and, with a capillary tube, picking up drops of urine formed at the nephropore. 
This method was satisfactory for getting urine samples for freezing point deter- 
mination, but not for finding rates of urine flow. The animals died after a short 
period under these conditions, presumably because the water clinging to the gills 
rapidly became anaerobic. 

To overcome this difficulty the following arrangement was devised. A short, 
bent piece of glass tubing was cemented to the bottom of a Petri dish, with one 
end opening horizontally and the other end vertically. The dish was filled with 
water of the desired concentration to a depth that covered the horizontal end of the 
tubing. The animal was removed from the adapting medium, grasped firmly by 
the coxal plate of the first right thoracic leg with watchmaker's forceps, and backed 
gently into the open end of the tube. The forceps was then clamped in place. 
The coxal plate is extremely thin, but is broad and usually is quite hard and sturdy. 
Grasping it had no apparent effect upon the rate of urine flow. Various diameters 
of glass tubing were used so that the animals of different sizes could be fitted 
snugly. Thus, although the animal was held firmly at only one point, its move- 
ment was greatly restricted and it could not rotate about that point. After the 
animal had been introduced into the tube and had quieted, the water surrounding 
the tube was removed and replaced with paraffin oil. Surface tension held the 
water in the tube during the change, which left the animal with most of its body 
in water and only its head protruding into the oil. The water was aerated by a 
bubbler made of fine tubing inserted into the upper end of the tube. 

Once the head and the nephrocones on the second antennae were surrounded 
by oil, it was possible to collect urine. For this purpose micropipettes were made 
from 0.3- or 0.5-mm. bore capillary tubing, tapered at the tip and graduated with 
strips of millimeter graph paper. The pipettes were filled with oil to avoid a strong 
capillary pull. They were then immersed in the oil bath with their tips capping 
the nephrocones. In some cases, especially with the smaller animals, the surface 
tension at the oil-urine interface of the minute nephridial opening apparently was 
sufficient to prevent the urine from flowing. To start the flow, it was usually 
necessary merely to touch the inside wall of the pipette to the nephropore and 
thereby break the interface. The rate of urine flow was determined from the progress 
of the urine drop up the pipette. Both nephridia were sampled simultaneously; 



OSMOTIC REGULATION IN GAMMARIDS 



229 



their rates rarely differed by as much as a factor of two. Urine flow was measured 
for about half an hour, a reading of the level in the pipettes being taken every five 
minutes in most cases, every two minutes in some others. The mean rates of urine 
flow for each side were added to give the total rate. 

One difficulty was encountered with this method. Although most animals pro- 
duced a measurable amount of urine when in any of the lower salinities, some did 
not. In most cases these animals had been injured in handling. Recently molted 
animals with soft exoskeletons were especially prone to damage. The results from 



1.6 



o 



1.4 - 



- 1.2 - 



Q 
O 
O 



CD 
U. 

O 



1.0 



P 0.8 

o: 

i- 
2 
UJ 

o 0.6 

O 
o 

o 

: 0.4 



IS) 

o 



0.2 



0.0 




o MARINOGAMMARUS FINMARCHICUS 

GAMMARUS OCEAN ICUS 
D GAMMARUS TIGRINUS 

GAMMARUS FASCIATUS 



0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 

OSMOTIC CONCENTRATION OF MEDIUM IN M/KG 



1.6 



FIGURE 1. Relationship of concentration of blood (in moles of ideal solute per kg. of 
water) to concentration of medium in Marinogainmarus finmarchicus and Gammarus oceanicus 
(marine), G. tigrinus (brackish-water), and G. jasciatus (fresh-water). Points represent 
means ; vertical bars represent 2 X the standard errors. 

damaged animals were discarded. There were a few cases of no urine flow for 
which there was no observed damage. Data from these animals are included inr 
the graphs but not in the statistical calculations. 

The choice of media was determined by the desire to find the relation of the 
rate of urine flow not only to the external concentration, but also to the osmotic 
gradient (obtained by subtracting the sea- water concentrations from the blood con- 
centrations in Figure 1). The media chosen were ones in which a wide range of 
gradients would be expected: 0.03, 0.2, 0.61, and 0.93 molal sea water for the 
marine species G. oceanicus, and fresh water, 0.2, and 0.61 molal sea water for 



230 



HENRY O. WERNTZ 



the fresh-water G. jasciatus. The experiments were carried out at approximately 
15 C. 

RESULTS 
Osmoregulatory behavior 

The four species regulated osmotically in the manner shown in Figure 1. Each 
point is the mean of determinations on several animals (see methods section). The 
vertical lines at each point give 2 X the standard error of the mean, or the 95% 
confidence interval for the mean. 



O GAMMARUS OCEANICUS 
GAMMARUS FASCIATUS 



O? 




0.2 0.4 0.6 0.8 1.0 

OSMOTIC CONCENTRATION OF MEDIUM, M/KG 

FIGURE 2. Relationship of rate of urine flow to external concentration in Gammarus 

oceanicus (marine) and G. jasciatus (fresh-water). Points represent individual animals. 

Points marked with a question mark are not included in the statistics. Curves connect the 
means. Vertical lines show 2 X the standard errors. 

The marine species (M. finmarchicus and G. oceanicus) exhibited nearly iden- 
tical relationships of blood concentration to external concentration. In full strength 
sea water their blood was nearly isotonic to the medium. They began to regulate 
the blood concentration when sea water was diluted only slightly to below 0.85 
molal. As the medium was further diluted, they maintained a progressively greater 
concentration gradient, and in 0.2 molal sea water the blood concentration was still 
0.7 molal. In the most dilute sea-water media the gradient failed to increase fur- 
ther, or even decreased, but all animals in 0.03 molal sea water survived. In fresh 
water all M. finmarchicus died, but three out of ten G. oceanicus survived 24 hours, 



OSMOTIC REGULATION IN GAMMARIDS 



231 



-or long enough to be sampled, although their condition was probahly moribund. 
In another experiment no G. oceanicus survived even this long in fresh water. 

The brackish-water G. tigrinus displayed an osmotic behavior intermediate be- 
tween that of G. oceanicus and M. finmarchicus on the one hand and that of G. 
jasciatus on the other. It regulated in media of less than 0.65 molal concentration 
and kept the blood concentration at 0.55 molal in 0.1 molal sea water and at 0.44 
molal in fresh water. The most remarkable aspect of this experiment is that the 



GAMMARUS OCEANICUS 
IN 0.03 M/l SEAWATER 
O IN 20 M/l SEAWATER 
a IN 0.6 I M/l SEAWATER 
A IN 0.93 M/l SEAWATER 



GAMMARUS FASCIATUS 
IN FRESHWATER 
IN 0.20M/I SEAWATER 
IN 0.61 M/l SEAWATER 




O.I 0.2 0.3 0.4 0.5 

OSMOTIC GRADIENT BETWEEN BLOOD AND MEDIUM, M/KG 

FIGURE 3. Relationship of rate of urine flow to osmotic gradient between blood and 
medium in G. oceanicus and G. fasciafus. Slope (S) determined by method of least squares. 

species survived after direct transfer both to fresh water and to 1.5 molal sea 
water, although its mortality in either of these extreme concentrations was about 
50%. In other experiments, some individuals survived a period of weeks in fresh 
water. 

G. fasciatus had a blood concentration of a little over 0.3 molal in fresh water, 
its normal habitat. When the animals were transferred into dilutions of sea water, 
the blood concentration increased only slightly and became isotonic in media with 
a concentration more than 0.5 molal. The species survived well in 0.6 molal sea 
water, but most individuals in 0.8 molal and all in 1.0 molal sea water died. 



232 

Urinary rate 



HENRY O. WERNTZ 



Rate of urine flow is plotted against concentration of medium in Figure 2. 
Despite considerable variability in the data, it is apparent that in both species the 
rate of flow was much greater in the more dilute media. Urine production was 





C 


.2 0.4 


6 0.8 1.0 


+ 0.1 


r~ 


~r~ i ~~i 


1 1 




0.31) (0.36) 


(0.65)1 


i 


o 

g-O.I 

O 

CD 


8 


GAMM 


ARUS FASCIATUS 


i-0.2 


I* 







LU 

z 

-03 


> 


A 





_ 
O 



O 

z 

O 
O 



UJ 

LU 



to. ir- 



m -O.I 




O 

H 

LU 

LU 
Ll_ 

U_ 



-0.2 



-0.3 



(0.76) 



(0.95) 



GAMMARUS OCEANICUS 



L 



_L 



0.2 0.4 0.6 0.8 

OSMOTIC CONCENTRATION OF MEDIUM, 



_J 
1.0 

M/KG 



FIGURE 4. Relationship of urine concentration to blood concentration in various media, 
in G. oceanicus and G. fasciatus. Filled circles represent the difference: (urine concentration 
minus blood concentration) for individuals. Open circles show the mean difference ; vertical 
lines 2 X the standard errors. Mean blood concentrations are given in parentheses. 

detectable at higher salinities in G. oceanicus than in G. fasciatus. Moreover, the 
maximum rate (over 4% of the body weight per hour) in G. oceanicus was more 
than twice as great as the maximum rate (around 2%} in G. fasciatus. 

A few determinations on specimens of both species at 25 C. indicated that the 
rate was roughly twice that of animals in the same concentrations at 15 C. 



OSMOTIC REGULATION IN GAMMARIDS 233 

The rate of urine flow in the two species is plotted against the osmotic gradient 
in Figure 3, and the regression lines are shown. The correlation coefficients for 
these data were 0.84 for G. oceanicus and 0.75 for G. fasciatus. In both cases the 
probability that a coefficient this size arose by chance is less than 0.001, as deter- 
mined by a t-test. The slopes or regression coefficients show that the best estimate 
of the rate of urine production was 10.5 % of the body weight per hour per molal 
gradient in G. oceanicus and 5.9% in G. fasciatus. If the regression coefficients 
are compared by a t-test presented by Fisher (1950), it is found that the probability 
they were drawn from the same statistical population is approximately 0.02. 

Urine concentration 

The difference between the urine concentration and the blood concentration for 
the two species of Gammarus in different salinities is shown in Figure 4. The urine 
of G. oceanicus was approximately isotonic to the blood in all media except the 
most dilute. In the latter, it was hypotonic by a small but statistically significant 
amount. The urine of G. fasciatus in fresh water was strongly hypotonic to the 
blood. In 0.2 molal sea water there was a great deal of variability, the urine 
ranging from equal in concentration to the blood to more dilute than the medium. 
The data for 0.6 molal sea water unfortunately are rather inadequate, partly be- 
cause of difficulty in getting enough urine for a freezing point sample and partly 
because one of the samples seems to have been contaminated. However, the urine 
at this salinity appeared to be isotonic with the blood. 

DISCUSSION 
Role of nephridium 

Since the rate of water movement across a semipermeable membrane is propor- 
tional to the osmotic gradient, the demonstration that the urinary rate is propor- 
tional to the concentration difference between blood and medium supports the 
hypothesis that the water which enters Gammarus by osmosis is eliminated as 
urine. However, in order to prove the hypothesis completely, it would be neces- 
sary to measure the osmotic uptake of water by some independent method, and to 
demonstrate that urine flow is adequate to account for the removal of this water. 
Although this plausible mechanism of water balance in crustaceans has often been 
proposed before, the evidence regarding both the proportionality of urine flow 
and its adequacy has not been conclusive. The evidence on each of these points 
will be considered in turn in the following paragraphs. 

In most studies of urinary rate in crustaceans and its relation to the external 
concentration, the method has been to plug the nephridial opening and to assume 
that subsequent weight increases represented normal urine formation. This pro- 
cedure is suspect a priori in an armored animal such as a crustacean. One would 
expect blocking the nephridial opening to cause an increase in the hydrostatic 
pressure in the nephridium and, if urine formation occurs through filtration, to 
cause a change in the rate of this process. Indeed, plugging the kidneys has been 
reported as leading to the death of the experimental animals by Herrmann (1931) 
in Astacns; and by Nagel (1934) in Carcinus. While this method has sometimes 
yielded results in accord with those from other methods (see discussion of Parry's 



234 



HENRY O. WERNTZ 



work, below), it has also given ambiguous results. For example, Nagel (1934) 
found that when crabs with plugged nephridia were transferred to brackish water 
they gained slightly more on the average than did similar crabs left in sea water. 
However, t-tests upon Nagel's data show the difference between transfer and 
control groups was of dubious statistical significance : p > 0.2 for the difference 
between rates for two groups given the same period of exposure ; p = 0.05 for 
the difference between rates for two groups given different exposure times. 

The most thorough study of the influence of salinity variation upon the rate of 
urine production in crustaceans was made by Parry (1955) in the prawn, Palae- 
monetes varians. This investigator estimated the rate of urine flow from four 
independent measurements: (1) the time for clearance of injected dye, (2) the 
frequency of micturition from a bladder of known size, (3) the weight change after 
blockage of the nephropores, and (4) the volume of urine collected by cannulation 
of one of the nephropores. All four methods gave similar results. Parry did not 
relate the rates to the gradient between blood and medium, since she did not 

TABLE I 

Relation between osmotic gradient and rate of urine flow in Palaemonetes varians 



Osmotic cone, of medium 
(moles/kg.) 


Gradient between blood and 
medium (calculated from 
Panikkar, 1941) 


Urine flow as % of body 
wt./hr. (determined by 
rate of micturition, 
Parry. 1955) 


Ratio: 
/ Urine flow\ 


\ gradient / 


.05 


.49 


1.63 


3.3 


.15 


.39 


1.06 


2.7 


.25 


.29 


.94 


3.2 


.50 


.07 


.15 


2.1 


.67 


-.10 


.45 


-4.5 


.85 


-.25 


.40 


-1.6 


1.00 


-.34 


.42 


-1.2 


1.20 


-.51 


.11 


.22 



determine blood concentrations. However, if we use concentrations determined 
for the same species by Panikkar (1941; also cited by Parry), we can consider 
the data from this viewpoint (Table I). We see that during hypertonic regulation 
in external concentrations of less than about 0.6 mole/kg., the rate of urine produc- 
tion was approximately proportional to the osmotic gradient. It is conceivable 
that the small increase in urinary rate during hypotonic regulation has the same 
cause as the increased urinary rate in injured marine teleosts (Smith, 1932), 
namely, the necessity of excreting the extra divalent ions in swallowed sea water. 
The rate of water entry into crustaceans has not been determined with sufficient 
precision to permit an exact comparison with the rate of urine flow. In the previ- 
ously mentioned study, Parry found that the half-time for penetration of heavy 
water into P. varians in nearly isotonic conditions was one-half to three-fourths 
hour. Such determinations may be used (cf. Lockwood, 1961) to predict the rate 
at which water should be absorbed under a specified osmotic gradient. Treating 
Parry's data in this manner, one may calculate that an amount equaling 1.9% to 
2.5% of the total body water (or 1.5% to 2.0% of body weight, if a prawn is 80% 
water ) should enter per hour and per gradient of one mole/kg. The values for 



OSMOTIC REGULATION IN GAMMARIDS 235 

urinary rates in column 4, Table I, are roughly comparable (2.1% to 3.3% of 
body weight per hour and per gradient of one mole/kg.). 

Lockwood (1961) has determined the rate of penetration of tritiated water 
into two British species of Gammarus: G. pulex and G. duebeni. The difference 
between the two species was not statistically significant. The half-time for pene- 
tration in G. duebeni was 13.9 minutes, which leads to an estimated rate of water 
entry of about 4.7% of body weight per hour per gradient of 1 mole/kg. While 
this is less than the rates of urine flow shown by the slopes in Figure 3, it is of the 
same order of magnitude. 

In summary, the similarity of observed urinary rates to those predicted from 
permeability studies is at least in accord with the view that osmotically absorbed 
water is eliminated as urine. The data are not sufficiently exact, however, to be 
decisive. 

Comparison of species 

A. Comparison of G. fasciatus and G. oceanicus 

Three differences have been shown which give the fresh-water G. fasciatits a 
lower rate of urinary salt loss than the marine G. oceanicus. They are (1) a 
smaller concentration gradient between blood and medium, (2) a lesser permeability 
to water, 4 and (3) a urine which is hypotonic to the blood. 

The first two differences result in a lower rate of urine flow ; the third results 
in a smaller salt loss for a given volume of urine. The quantitative result of these 
differences is shown in Figure 5, in which the urinary rate is multiplied by the urine 
concentration to give the rate of salt loss in micromoles of solute particles per hour 
in a 100-mg. animal. 

B. Comparison with other species 

None of the species in this study was the same as any in the study by Beadle 
and Cragg (1940a). Two of their species are not known to occur at all in 
North America, and the other two do not occur in southern New England, where 
the present study was made. However, Beadle and Cragg's findings were similar 
to those presented here in that their more marine species, G. locusta and G. 
(Marinogammarus) obtusatus. regulated the blood concentration at a high level ; 
the brackish-water G. duebeni regulated at an intermediate level ; and the fresh- 
water G. pulex, at a low level. The major differences from the present study were 
that their marine species, especially G. obtusatus, did not survive in nearly as dilute 
sea water as did G. oceanicus and M. finmarchicus ; and that none of their species 
could survive in both fresh water and full strength sea water. (Later, however 
[1940b], Beadle and Cragg reported on a fresh- water race of G. duebeni.') 

The recent papers of Lockwood (1961) and of Shaw and Sutcliffe (1961) 
were concerned largely with water entry, which has already been discussed, and 

4 From the difference in slopes in Figure 3. The rates of urine flow in this figure are 
based on the weights of the animal, whereas permeabilities should be based on the surface 
areas, which are not known. It can easily be shown, however, that since G. oceanicus averaged 
about five times the weight of G. fasciatus, conversion of urinary rate to permeability would 
increase the difference found between the species. 



236 



HENRY O. WERNTZ 



with sodium uptake, which will be discussed in the next section. Several aspects 
of the salt loss remain to be compared. Shaw and Sutcliffe found that the rate 
of salt loss by all routes from a 40-mg. G. duebeni varied from 0.17 to 0.76 micro- 
mole of NaCl per hour, being less when the animal had been adapted to a more 
dilute medium. For G. pule.v, the rate was 0.09 to 0.18 rnicromole per hour. 




0- 2 



0-4 



0. 6 



0.8 



1.0 



CONC. OF MEDIUM , M/KG 



FIGURE 5. Comparison of rate of urinary salt loss in G. occanicus and G. fasciatus. Points 
represent individuals. Curves connect means. Vertical lines show 2 X the standard errors. 

Lockwood determined that in very dilute sea water both the brackish-water G. 
duebeni and the fresh- water G. pule.v formed urine hypotonic to the blood. In the 
higher concentrations of sea water the urine of G. duebeni became isotonic to the 
blood, but the urine of G. pulex remained hypotonic to the blood and also to the 
medium. Estimating the urine flow from the penetration of tritiated water, Lock- 
wood calculated that only 0.10 rnicromole per hour of the NaCl loss from G. duebeni 



OSMOTIC REGULATION IN GAMMARIDS 237 

could be accounted for by the urine. Apparently there is a substantial loss by some 
other route, probably across the body surface. 

From Figure 5 one can calculate that a 40-mg. G. oceanicus in the most dilute 
sea water would retain salt less efficiently than G. duebeni and would lose about 
0.34 micromole of NaCl per hour through the urine alone. A G. fasciatus of the 
same size in fresh water would lose only about 0.02 micromole per hour. 

Mechanism of salt uptake 

The differences demonstrated between G. fasciatus and G. oceanicus bring to 
the fore a curious problem. The steady-state osmotic gradient maintained by these 
animals represents the conditions at which salt loss by all routes just equals salt 
uptake. Since the urine concentration and flow rate are more favorable for osmotic 
regulation in G. fasciatus than in G. oceanicus, how can the latter maintain a much 
greater osmotic gradient when both species are in the same medium? One logical 
possibility is that G. fasciatus is more permeable to salt than G. oceanicus and loses 
more salt by diffusion. This would be contrary to the results of other investigations 
of the comparative permeabilities of marine, brackish-water, and fresh-water crus- 
taceans (Gross, 1957; Bethe, 1930; Nagel, 1934). The only alternative explana- 
tion of the apparent paradox is that when both species are in the same medium, 
the rate of salt uptake is greater in G. oceanicus. 

From the steady-state condition it is apparent that when a gammarid is in a 
high external concentration, the mechanism for salt uptake must be operating either 
not at all or at only a very low rate, since the gradient is zero. The gradient be- 
comes appreciable that is, uptake begins only when the concentration is brought 
below a certain critical level. Activation of the mechanism must be gradual, since 
the osmotic gradient increases gradually as concentration is further lowered. The 
critical concentration at which uptake and regulation begin is characteristic of the 
species, being highest in the marine species (G. oceanicus and M. finmarchicits] , 
next highest in the brackish-water species (G. tigrinus*), and lowest in the fresh- 
water species (G. fasciatus'). Consequently, in most salinities the uptake mecha- 
nism is more completely activated in G. oceanicus than in G. fasciatus and the former 
species maintains a greater gradient. The degree of activation of the uptake mecha- 
nism might depend upon either the concentration of the medium or that of the 
blood. The latter alternative is favored by Shaw's (1959) demonstration that in 
the crayfish, sodium uptake from fresh water increases with decreasing blood 
concentration. 

In most dilute media the gradient maintained by a species falls below the maxi- 
mum, indicating a drop in the salt uptake rate. In this circumstance G. oceanicus 
no longer has a gradient greater than G. fasciatus; indeed, in fresh water it cannot 
maintain an internal concentration sufficiently high for survival. 

Shaw and Sutcliffe (1961) directly measured uptake by Gammarus from very 
dilute media and found that with increasing external concentration the uptake rate 
rose asymptotically to a maximum. Since this behavior can be described by the 
Michaelis equation for effect of substrate concentration on rate of an enzyme- 
intermediated reaction, the implication is that in very dilute media the uptake 
mechanism becomes unsaturated. Using this interpretation, Shaw and Sutcliffe 
further concluded that the uptake mechanism of G. pule.v had a greater affinity for 



HENRY O. WERNTZ 

ions than that of G. duebeni, and therefore the former could maintain a greater 
gradient in the most dilute media. This analysis would equally well explain the 
difference in osmotic behavior between G. jasciatus and G. oceanicus. 

SUMMARY 

1. The relationship of the osmotic concentration of the blood to that of the 
external medium is described for four species of gammaricl. In media of various 
concentrations each species regulates its blood concentration in a manner that reflects 
its natural habitat. The marine species, Marinogainniarus finmarchicus and Gam- 
inarus oceanicus, regulate their blood concentration at the highest level ; the brack- 
ish-water species, G. tigrinus, regulates at a lower level ; and the fresh-water species, 
G. jasciatus, regulates at the lowest level. Moreover, M. finmarchicus and G. 
oceanicus die in fresh water; G. jasciatus dies in full-strength sea water; but G. 
tigrinus survives both in fresh water and in sea water up to at least 1.5 molal. 

2. The rate of urine production in G. jasciatus and G. oceanicus is proportional 
to the osmotic gradient between blood and medium, indicating that urine formation 
represents elimination of osmotically absorbed water. The coefficient of propor- 
tionality is smaller in G. jasciatus (5.9% of body weight per hour per molal 
gradient) than in G. oceanicus (10.5%}, indicating that the latter species is more 
permeable to water. 

3. The urine of G. oceanicus is nearly isotonic to the blood in all media. The 
urine of G. jasciatus is much more dilute than the blood. 

4. The differences in flow and in concentration of urine combine to give 
G. oceanicus a much greater rate of urinary salt loss than G. jasciatus. 

5. The osmotic gradient maintained by each species varies in a way that indi- 
cates the animals have an ionic uptake mechanism which is gradually activated as 
the salinity is lowered. In all except the most dilute media, it appears that the 
mechanism is more completely activated and takes up salt more rapidly in G. oceani- 
cus than in G. jasciatus. 

LITERATURE CITED 

BAUMBERGER, J. P., AND J. M. D. OLMSTED, 1928. Changes in the osmotic pressure and water 

content of crabs during the molt cycle. Physiol. Zool., 1 : 531-545. 
BEADLE, L. C., AND J. B. CRAGG, 1940a. Studies on adaptation to salinity in Gammams sp. 

I. Regulation of blood and tissues and the problem of adaptation to fresh water. 

/. Exp. Biol, 17: 153-163. 
BEADLE, L. C., AND J. B. CRAGG, 1940b. Osmotic regulation in freshwater animals. Nature, 

146: 588. 
BETHE, A., 1930. The permeability of the surface of marine animals. /. Gen. Physiol., 13 : 

437_444. 

BOUSFIELD, E. L., 1958. Fresh-water amphipod crustaceans of glaciated North America. 
Canad. Field-Nat., 72: 55-113. 

BURIAN, R., AND A. MUTH, 1924. Die Exkretion (Crustaceen). In: Winterstein's "Hand- 
buch der vergl. Physiol." Bd. 2, 2. Halfte, pp. 633-695. 

FISHER, R. A., 1950. Statistical Methods for Research Workers, llth ed. Hafner Publ. Co., 
New York. 

GROSS, WARREN J., 1957. An analysis of response to osmotic stress in selected decapod crus- 
taceans. Biol. Bull, 112: 43-62. 

HERRMANN, F., 1931. Uber den Wasserhaushalt des Flusskrebses (Potamobius astacus Leach). 
Zeitschr. vergl. Physiol., 14 : 479-524. 



OSMOTIC REGULATION IN GAMMARIDS 239 

HUF, E., 1936. Der Einfluss des mechanischen Innendrucks auf die Flussigkeitsausscheidung 

bei gepanzerten Susswasser- und Meereskrebsen. Pflug. Arch. ges. Phvsiol.. 237: 

240-250. 
KINNE, O., 1952. Ein neues Gerat zur Bestimmung der Gefrierpunktserniedrigung kleiner 

Fliissigkeitsmengen. Vcroff. Inst. Meeresforsch., Bremerhaven., 1 : 47-51. 
KINNE, O., 1954. Die Gammarus-Arten der Kieler Bucht. Zool. Jahrb., Abt. Sysl., Oknl, u. 

Geogr. Ticrc, 82 : 405-424. 
LOCKWOOD, A. P. M., 1961. The urine of Gammarus ducbcni and G. pulex. J. Exp. Biol., 38 : 

647-658. 
NAGEL, H., 1934. Die Aufgaben der Exkretionsorgane und der Kiemen bei der Osmoregulation 

von Car dims macnas. Zeitschr. vergl. Physiol., 21 : 468-491. 
PANIKKAR, N. K., 1941. Osmoregulation in some palaemonid prawns. /. Mar. Biol. Assoc., 

25: 317-359. 
PARRY, G., 1955. Urine production by the antennal glands of Palaemonetes varians (Leach). 

/. Exp. Biol., 32 : 408-422. 
RAMSAY, J. A., 1949. A new method of freezing point determination for small quantities. 

/. Exp. Biol, 26: 57-64. 
SCHELLENBERG, A., 1937. Schliissel und Diagnosen der dem Susswasser-Gammarus nahestehen- 

den Einheiten ausschliesslich der Arten des Baikalsees und Australiens. Zool. 

Ans., 117: 267-280. 

SCHWABE, E., 1933. tlber die Osmoregulation verschiedener Krebse (Malacostracen). Zeit- 
schr. vergl. Physiol., 19: 183-236. 
SEGERSTRALE, S. G., 1947. New observations on the distribution and morphology of Gammarus 

saddachi Sexton, with notes on related species. /. Mar. Biol. Assoc., 27 : 219-244. 
SEXTON, E. W., 1939. On a new species of Gammarus (G. tigrinus} from Droitwich district. 

/. Mar. Biol. Assoc., 23 : 543-552. 
SEXTON, E. W., AND G. M. SPOONER, 1940. An account of Marino gammarus (Schellenberg) 

gen. nov., with a description of a new species, M. pirloti. J. Mar. Biol. Assoc., 24 : 

633-682. 
SHAW, J., 1959. The absorption of sodium ions by the crayfish, Astacus palllpes Lereboullet. 

I. The effect of external and internal sodium concentrations. /. Exp. Biol, 36 : 

126-144. 
SHAW, J., AND D. W. SUTCLIFFE, 1961. Studies on sodium balance in Gammarus ducbcni Lillje- 

borg and G. pulex pulex (L.). /. Exp. Biol, 38: 1-15. 
SMITH, HOMER W., 1932. Water regulation and its evolution in fishes. Quart. Rev. Biol, 

7: 1-26. 






Vol. 124, No. 3 June, 1963 

THE 

BIOLOGICAL BULLETIN 

PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY 




THE SIGNIFICANCE OF THE CAUDAL EPIDERMIS IN 
ASCI DI AN METAMORPHOSIS l 

RICHARD A. CLONEY 
Department of Zoolo</y, I'uircrsity of Washington, Seattle 5, U't/shini/ton 

Free swimming ascidian larvae normally attach to solid substrata before the 
onset of metamorphosis. Morphogenetic changes include tail resorption, reorienta- 
tion of the digestive system, heart and adult neural tissues (or their primordia) and 
the formation of epidermal ampullae. Degeneration of the larval nervous and 
sensory systems and the loss of the outer layer of tunic follow rapidly. 

The most astonishing event in metamorphosis is the withdrawal of the tail. In 
the simple ascidians, Boltcnia rillosa, Pyiira hanstor, Stydct ijihbsii, and Stychi 
partita, the major part of tail resorption is completed within 10 minutes. In the 
compound ascidian Amaroucium constellatum the tail is completely withdrawn into 
the posterior region of the trunk in about 6 minutes. Histological studies of 
Boltcnia, Pyiira and Stycla (Cloney, 1961 a) have demonstrated that a disruption 
of the intercellular cementing substances or binding forces between the notochordal 
and muscle cells occurs during tail resorption. These disruptive changes begin 
proximally and progress distally as the tail shortens. The anterior end of the 
notochordal sheath first explosively ruptures. This allows the notochordal cells and 
the extracellular matrix of the notochord to flow into the posterior region of the 
trunk. Simultaneously the muscle cells buckle and their myofibrils become disar- 
ranged and are no longer oriented parallel to the axis of shortening. The nerve 
cord and endodermal strand are passive and become compressed as the tail shortens. 
The epidermis thickens but shows no signs of dissociating into separate cells. This 
pattern of tail resorption will be referred to as Type 1. 

In dona intcstinalis (Weiss, 1928), PJiallusia niannnillata, Clarclina lepidi- 
fonnis (Berrill, 1947), Auuiroiicinin constellatum (Scott, 1952), Ascidia nitjra 
(Grave, 1935) and Ascidia callosa (Cloney, 1961a), the notochorcl-muscle-nerve 
cord complex (NMN-complex) remains intact as a unit as the caudal tissues are 
withdrawn. The NMN-complex separates from the epidermis and moves into the 

1 This investigation was supported in part by Public Health Service Research Grant 
RG-9936 from the Division of General Medical Sciences. 

241 
Copyright 1963, by the Marine Biological Laboratory 



242 RICHARD A. CLONEY 

posterior region of the trunk. In some species the complex is formed into a helix 
as it moves. The epidermis finally forms a cap over the resorhed tissues. This 
will be called Type 2 tail resorption. 

When tails of Boltcnia larvae were excised before the onset of metamorphosis, 
the distal fragments remained alive and were observed to twitch or even swim 
about for one or two days after the operation, but they showed no signs of the histo- 
logical changes associated with tail resorption. Connection with the trunk is 
normally essential for these changes. 

If, however, the tails were excised through the proximal region of shortening 
after the onset of metamorphosis they could undergo shortening in isolation. Evi- 
dently the mechanism for tail resorption resides within the tail tissues after the 
process has begun. Isolated tail can also be induced to shorten with proteolytic 
enzymes. 

Instructive cases of partial tail resorption in chloretone-treated Ciona larvae 
have been observed by Weiss 1928). In these specimens the NMN-complex re- 
mained rigid but the epidermis sometimes became torn or ruptured in places and 
pulled into a mass near the tip of the tail. Weiss emphasized that the epidermis 
can therefore undergo typical regressive changes by itself. Direct observations of 
tail resorption, however, led him to conclude that both epidermal and tonic muscu- 
lar contraction are probably essential for complete tail involution in Ciona. 

Berrill (1947) contended that tail resorption is caused principally by the shrink- 
age of the epidermis due to its so-called "nutritional exhaustion" in Ciona, Asci- 
dicl/a, PJiallusia, Stycla, Styclopsis, Distoinus, Stolonica, Clai'dina and Distaplia. 
This opinion was supported by Scott (1952). Cloney (1961 a) argued that the 
epidermis is an active tissue, as evidenced by ( 1 ) the resorption of the epidermal 
adhesive papillae within two to three minutes after the onset of metamorphosis in 
Boltcnia, Pyitra, Stycla and Ascidia; (2) the formation of an invagination of the 
epidermis behind the resorbed tail tissue elements at the end of tail resorption, and 
(3) the rapid formation of the epidermal ampullae which spread out over the sub- 
stratum shortly after attachment of the larva. There is no evidence of "nutritional 
exhaustion" in these species ; indeed, the epidermal cells display considerable ac- 
tivity and they also contain yolk granules. (For a fuller discussion of this point 
see Cloney, 1961a. ) 

It was suggested that the release of a proteolytic enzyme in the posterior region 
of the trunk in Boltcnia could account for the rupture of the notochord, the dissoci- 
ation of the muscle and notochordal cells, and that the epidermis contracts actively 
as a unit, forcing the dissociated tissues into the trunk. 

But, since in Boltcnia, P\<nra, and Stycla, it was not possible to isolate or to 
observe the independent contraction of the tail epidermis, only indirect evidence for 
its active role in the resorption of the tail could lie obtained. Consideration of 
Type 1 and Type 2 tail resorption led to the following questions : ( 1 ) Does con- 
traction of the muscle cells play any part in the process? All analyses of tail re- 
sorption have shown that the epidermis is important, but (2) what is the mech- 
anism of epidermal contraction? (3) What factors initiate and synchronize the 
observed histological changes with general metamorphosis? (4) Can the two dis- 
tinct types of tail resorption be explained with a single hypothesis ? This report 
on the compound ascidian, Ainaroiiciiiin constcllatuin, provides direct evidence for: 



ASCIDIAN METAMOKI'HOSIS 243 

(1) the active contraction ot the epidermis in unaesthetized larvae, (2) the insig- 
nificance of the muscle cells in the overall mechanism of tail resorption, (3) a pos- 
sible explanation for other historical changes in the caudal tissues of ascidians 
during tail resorption. 

METHODS 

Observations were made on larvae of the compound ascidian Amaroucium con- 
stcllatiiiii collected by the Supply Department in the vicinity of the Marine Bio- 
logical Laboratory, Woods Hole, Massachusetts. Colonies were kept in running 
sea water in a light-tight box for 8 to 18 hours prior to the period when they were 
needed. Colonies were then exposed to light in a small dish of sea water. This 
method, used by Scott (1952), induces the release of swimming larvae in about 20 
minutes. Observations of living animals were made with and without supported 
coverslips under a compound microscope equipped with NA 0.25 and 0.65 objec- 
tives. In some cases metamorphosis was stimulated with a 1:500,000 dilution of 
Janus green B. Cinematographic records were made of both normal tail resorption 
and the effects of experimental interference with tail resorption. Tissues were 
fixed in 2.7% OsO 4 buffered in 0.1 M vS-collidine and embedded in Epon, accord- 
ing to the method of Luft (1961). Sections were cut at \ /JL and stained with 
Richardson's stain (Richardson ct a/., 1960). Photomicrographs of sections were 
made with a Zeiss NA 1.25 planachromat objective. 

RESULTS 
A. Structure of the larral tail 

Details of the larval anatomy have been described by Grave (1921 ) and Scott 
(1946). The following descriptions of the tail tissues are limited to details regarded 
as essential to a discussion of tail resorption. 

Tunic. The entire trunk and tail of the larva is covered by two tough mem- 
branous layers of tunic. The outer layer forms the dorsal and ventral fins and is 
lost as a molt during metamorphosis. The inner layer is retained as part of the 
post-larval tunic. Free amoeboid cells are frequently found between the layers 
(Fig. 1). The fine structure of this complex structure will be described in a 
subsequent paper. 

Epidermis. The epidermis is a simple squamous epithelium supported basally 
by a thick amorphous basement membrane which lies in close contact with the sur- 
faces of the underlying muscle cells. In transverse sections of the tail the epi- 
dermis consequently tends to conform to the scalloped contours of the muscle bands 
(Fig. 1). 

Muscle. Four rows of muscle cells are arranged in bands on the right and left 
sides of the tail. (Since the tail is twisted 90 to the left during development, the 
top of the tail is considered to represent the anatomical right side and the bottom 
the left side. The dorsal and ventral fins lie in the frontal plane. ) Within each 
row, the muscle cells abut against each other at their ends without the intervention 
of a connective tissue septum. Kach muscle cell is roughly cylindrical in shape. 
Contractile myofibrils lie immediately beneath the sarcolemma in a single layer. 
They spiral along the course of the cell at an angle of about 18 to the right ( Grave, 
1921 ). They are disposed entirely around the periphery of each cell except for 



244 



RICHARD A. CLONEY 



NV 




MC 



FIGURE 1. Amaroucium constellatum; transverse section of the larval tail. The tail is 
covered by two layers of tunic ( T ) . The outer layer forms the dorsal and ventral fins. The 
basement membrane of the thin epidermis (E) lies in close contact with 8 rows of muscle cells 
(MC). Myrofibrils (my) lie in the periphery of each cell. They are underlain by a thick 
layer of mitochondria (mt). The muscle cells attach to a fibrous sheath of the notochord. 
Notochordal cells (NC) form a squamous epithelium beneath the sheath. The axis or lumen 
( 1 ) of the notochord is filled with a clear matrix. The nerve cord (NV) is visible on the 
right. One-micron Epon section ; Richardson's stain. 

small gaps where adjacent muscle cells lie in close contact. Contrary to the 
opinion of Grave (1921), Scott (1946) and Berrill (1947), the myofibrils termi- 
nate near the sarcolemma at the ends of each cell (Jackson, 1958; Cloney, un- 
published results). A dense layer of mitochondria is located beneath the myofibrils. 
The nucleus is located near the center of each cell and is surrounded by relatively 
clear cytoplasm. Lead acetate staining of ultrathin sections suggests the presence 
of glycogen in these areas. Large, irregular, dense bodies are often found in the 
cytoplasm of the muscle cells. In electron micrographs these bodies can be seen to 
contain myelinic figures. The band of muscle cells on the right side of the tail is 
often shifted slightly with respect to the left band of muscle, giving the tail an 
asymmetric appearance in section (Fig. 1 ). 

Notochord. The notochord forms the axis of the tail. In transverse sections 
the notochord appears elongate in the dorsal-ventral axis and somewhat irregular 
in shape. In electron micrographs, it is seen to be surrounded by an acellular 
filamentous sheath. The cells of the notochord are arranged in an epithelium 
and are attached to the inside of the notochordal sheath. The cells contain both 
proteid and lipid yolk granules. The axis or lumen of the notochord is filled with 
a clear matrix. The matrix stains poorly and has low electron density. Processes 
frequently extend from the notochordal cells into the lumen (Figs. 1 and 2). 

Nerve cord. The dorsal nerve cord has a well formed lumen in many places. 
The cord extends posteriorly from the visceral ganglion into the dorsal region of 



ASC1DIAN METAMORPHOSIS 



245 



the 1 tail. Nerve processes have not vet been observed passing from the cord to the 
muscle cells ( Fig. 1 ). 

B. Tail resolution 

The basic morphological changes accompanying metamorphosis in Ainarouciuni 
have been described by Scott (1952). Only the details of tail resorption will be 
considered here. 

In 14 recorded cases, the first detectable morphological changes associated with 
tail resorption occurred on an average of two minutes, two seconds after the 
discharge of the larval adhesive papillae (see Table I). The discharge of a sticky 
substance by the papillae is typically associated with attachment in Ainarouciuui, but 
metamorphosis may ensue with or without attachment. Metamorphosis may be 
spontaneous or it may be induced by a heterogeneous variety of substances (Lynch, 
1961 ). Changes in the tail tissues begin with the lifting away of the epidermis 
from the underlying notochord-muscle-nerve cord (NMN ) complex with the forma- 
tion of a subepidermal fluid-filled space (Figs. 3, 6). Simultaneously the notochord 
loses its turgidity. This results in a partial collapse of the tail (Fig. 3). The noto- 
chordal cells begin to round up and large vesicles sometimes become visible within 
the matrix of the notochord ( Fig. 6 ) . Under favorable optical conditions these 
vesicles can sometimes be seen to flow tow r ard the trunk. The entire NMN-com- 



TABLE I 

Timing of tail resorption. 

The first sign of the onset of metamorphosis in A. constellatuin is the discharge of the adhesive 
papillae. In column I, tabulations indicate the elapsed time in minutes and seconds from the discharge 
of the adhesive papillae to the first visible changes in the tail. Coin inn II indicates the time elapsed 
from the earliest changes in the tail to the completion of tail resorption. Column III indicates the 
total elapsed time from the discharge of the adhesive papillae to the completion of tail resorption. 
Measurements were made at 24.5C. Tabulations of 14 separate cases were arranged in order of 
increasing total time. 



Specimen 


Discharge of papillae 
to beginning of 
tail resorption 

I 


Beginning to completion of 
tail resorption 

II 


Total time 
III 


1 


2 '20" 


4'00" 


6'20" 


2 


I'lO" 


5 '40" 


6'50" 


3 


2'15" 


5 '00" 


7'15" 


4 


1'56" 


5 '3 7" 


7'33" 


5 


1'50" 


5 '45" 


7'35" 


6 


2'25" 


S'10" 


7'35" 


7 


1'25" 


6'20" 


7'45" 


8 


2'42" 


5 '05" 


7'47" 


9 


2 '30" 


6 '00" 


8'30" 


10 


2'45" 


5'50" 


8'35" 


11 


1'45" 


6'50" 


8'35" 


12 


2'05" 


6'32" 


8'37" 


13 


2'00" 


6'45" 


8'45" 


14 


1'21" 


7 '53" 


9'14" 


Average 


2'02" 


5'54" 


7 '55" 



246 



RICHARD A. CLONEY 




ASCIDIAN METAMORPHOSIS 247 

plex beneath the epidermis begins to buckle and fold as it moves into the posterior 
region of the trunk (Figs. 3, 4, 5). The epidermis continuously thickens and ap- 
pears to be under tension. The epidermis finally forms into a thick cap at the 
posterior end of the trunk, enclosing the entire XAIX-complex within the body 
cavity. 

As the tail tissues are withdrawn, the double-layered tunic which covers the 
tail becomes greatly folded. Near the completion of tail resorption the outer layer 
of tunic springs away from the underlying tissues and is pushed out, forming an 
empty sac. This outer layer drops away within a few hours as a cuticular molt. 
This outer layer need not be regarded as important to tail resorption because it can 
be pulled off with a pair of forceps before metamorphosis begins without affecting 
the process in any way. 

In observing tail resorption, one might infer that the epidermis is actively contract- 
ing because it appears to be under tension during this period, but it is of course im- 
possible to be certain of this from microscopic observations alone. 

To test the theory of active epidermal contraction, the tails of larvae which had 
just begun to metamorphose (after the epidermis has separated from the under- 
lying NMN-complex ) were either excised about halfway along their length or 
they were regionally damaged by touching them with a needle. When this was 
done the epidermis immediately split into tw r o units : an isolated distal piece and 
a proximal piece continuous with the trunk epidermis. Immediately following this 
operation both the proximal and the distal pieces of epidermis began to shorten 
over tlie surface of the underlying NAIN-comple.r (Figs. 7-10). Sometimes the 
epidermis contracted, pulling the underlying tissues with it for a short distance, and 
then the NMN-comp/c.v, evidently under tension, broke loose and straightened out 
again while the epidermis continued to contract. The proximal epidermis shortened 
into a thick annular ring at the base of the tail while the distal fragment pulled 
distally, forming a thickened cap of epithelium around the distal segment of the 
NMN-complex. This experiment was repeated several dozen times with the 
same results. 

In this experiment the epidermis manifests its capacity to shorten independently 

FIGURE 2. Amaroucium constcllatuin; right side of living larva. The epidermis (E), 
muscle cells (MO, notochordal cells (NO, notochordal lumen (1) and the tunic (T) are 
visible. Note the squamous epithelium within the notochord. Arrows indicate the position of 
the epidermis within the trunk. 

FIGURE 3. Amaroucium constcllatuin; right side of same specimen as Figure 2, about one 
minute after the onset of tail resorption. Note the appearance of a space (ss) beneath the 
epidermis and the marked change in the arrangement and shape of the notochordal cells (NO. 
The epidermis (E) has separated from the surface of the muscle cells (MO and has begun 
to thicken. The NMN-complex has begun the fold as a unit without a breakdown of the 
binding force or cementing substances between cells. 

FIGURE 4. Amaroucium constcllutuin ; right side of another specimen about two minutes after 
the onset of tail resorption. The NMN-complex has become more folded and has been partially 
forced into the posterior end of the trunk. The epidermis (E) has become thicker than it was 
in the larva. 

FIGURE 5. Amaroucium conxtclhititin ; right side of same specimen as seen in Figure 4 
about three minutes after the onset of metamorphosis. More than half of the tail is coiled 
within the posterior end of the trunk. The epidermis is greatly thickened. 



248 



RICHARD A. CLONEY 




mt 






FIGURE 6. Amarouciiim constcllatnin. Transverse section of the tail about two minutes 
after the onset of metamorphosis. A subepidermal space (ss) is prominent. The epidermis 
(E) is thicker than in the larva. Marked changes are apparent in the notochord. The cells 
(NC) tend to round up, filling the lumen. Many large vesicles are formed. The muscle cells 
(MC) increase somewhat in diameter as the NMN-complex bends and folds. One-micron 
Epon section ; Richardson's stain. 

of the other tissues. The NMN-complex is left hare by this action (except 
for the tunic which covers the entire organism). Under these circumstances the 
NMN-complex is never withdrawn into the trunk ; the muscle cells do not mani- 
fest any capacity to shorten. These tissues remain unresorhed while the rest of the 
larva completes its metamorphosis. 

Tails excised hefore the onset of metamorphoses do not undergo any of the 
characteristic histological changes ohserved in normal metamorphosing larvae or in 
tails excised after the onset of tail resorption. They twitch for many hours hut 
eventually degenerate. The tail tissues are not necessary for post-larval develop- 
ment in Ainarouciuni (Scott, 1952), or in Boltcnia ( Cloney, 1961b). The caudal 
tissues have no known prospective significance hut probably serve a nutritive 
function. 

This rupturing of the caudal epidermis occurs spontaneously in some larvae col- 
lected in culture dishes. Numerous cases have been ohserved in which meta- 
morphosis proceeded in the trunk without the resorption of the tail. This has also 
been reported by Scott (1952). C'lose inspection of more than a dozen of these 
specimens which failed to resorb their tails revealed that in all cases, there was a 
mass of epidermis at the base of the tail and at the tip of the tail, while the central 
portion of the tail was not covered by epidermal tissue. 



ASC1DIAN METAMORPHOSIS 249 

Failure to resorb tbe tail may be attributed, at least in these cases, to the 
rupture of the epidermal envelope which normally contracts and is essential to the 
withdrawal of the NMN -complex. 

The question of the energetics of contraction remains to be considered. The 
epidermis might shorten through elastic properties or it could actively contract and 
be dependent on aerobic oxidative processes for energy. The latter is suggested 
by the following experiment. 

After the beginning of metamorphosis and following the onset of tail resorption, 
when the epidermis of the tail has separated from the underlying muscle, the larvae 
were placed for varying periods in a solution of 5 X 10~ a to 1 : : 10~- M KCN in 
sea water regulated to pH 8.0 with HC1 (Fig. 11). Within about one minute 
after exposure to KCN the rate of shortening was slowed down. If the specimens 
were then washed in sea water, tail resorption would resume after a minute or two. 
Tail resorption is thus reversibly inhibited by potassium cyanide. Sodium azidr 
has a similar inhibitory effect in a concentration of 10~- M. Both of these sub- 
stances have been reported to reversibly inhibit the onset of metamorphosis in 
Ainaroiiciiini larvae (Lynch, 1961). Potassium cyanide and sodium azide are well 
known for their inhibitory affects on cytochrome oxidase, the terminal enzyme com- 
plex in the respiratory chain (Wainio and Cooperstein, 1956; Pearse, 1960). 

DISCUSSION 

The overall aspects of tail resorption in Aniaroncinui correspond to Type 2 as 
mentioned in the introduction. This pattern of tail resorption has features in 
common with Type 1. 

In both forms there are changes in the notochord. The notochordal matrix 
which contributes to the rigidity of the tail (hydrostatic skeleton) is released and 
leads to a loss of turgor and a partial collapse in both forms. This aspect of tail 
resorption was not reported by Weiss (1928) in his study of dona. In Boltenia, 
Styela and Pvura. both the muscle cells and notochordal cells become detached 
from the notochordal sheath as the tail is withdrawn. This process does not 
occur in Ainaronciitin, or other Type 2 forms. The muscle cells and notochordal 
cells remain associated in the same linear relationships, attached to the noto- 
chordal sheath. But in Amarouc'uun there is a rapid dissociation of the epidermis 
from the underlying muscle cells, and rapid changes in the notochordal cells. 

The activity of a proteolytic enzyme might account for changes in the noto- 
chord and a breakdown of adhesion between tissues in all species. Conklin (1931 ) 
has made a similar suggestion to explain tail resorption in StyeJa partita but no 
direct evidence for the existence of a proteolytic or hydrolytic enzyme that is re- 
leased or is activated at the time of metamorphosis has yet been found. 

Tt seems reasonable, in view of the experiments with Atnaroucium and the evi- 
dence of epidermal activities in Ciona (Weiss, 1928), and Boltenia. to generalize 
the statement that the contraction of the epidermis is responsible for forcing the 
tissues of the tail into the trunk at the time of tail resorption in ascidians. 

Weiss' contention that tonic muscular contraction must contribute to the with- 
drawal of the tail seems untenable for the following reasons: ( 1 ) In Type 1 tail 
resorption the muscle cells become dissociated, and histological analyses show that 
the contractile elements become disarranged early in tail resorption. (2) In Type 
2 tail resorption excision of the tail after the beginning of tail resorption leads to 



250 



RICHARD A. CLONEY 




ASCIDIAN METAMORPHOSIS 251 






the retraction of the epidermis alone, while the NMN-complex actually pushes 
away from the trunk as if released from tension as soon as it detaches from the 
epidermis. 

The mechanism of epidermal contraction is of general biological interest. The 
following is quoted from Hoffmann-Berling (1960, p. 346). 

"All in all, it may be stated that systematic comparisons have not yet uncovered 
anv dissimilarity which would indicate a fundamental difference between the con- 

, * 

traction of a muscle and the contraction of an undifferentiated cell. The differ- 
ences are only quantitative. The mechanism of muscle contraction is already 
present in the final form at the developmental stage of the single cell organism prior 
to the start of organ formation and tissue specialization. It is older than muscle 
itself." 

Unpublished electron micrographs revealed filaments in the epidermis of 
Amaroucium larvae, but filaments are commonly found in epidermal cells, as well 
as many other epithelial cells in which active contraction is at least not obvious. 
Contractility may be one of the fundamental properties of all cells but in the case 
of tail resorption, some cells contract in a specific and predictable way while other 
cells of the tail do not perceptibly contract. These epithelial cells may be suitable 
subjects for further investigations of this long-standing problem. 

SUMMARY 

1. Tail resorption in Ainaroitchiui is a rapid morphogenetic process. It is 
usually complete within only 6 minutes. 

2. The initiation of tail resorption is signaled by a rapid separation of the epi- 

FIGURES 7-10. Amaroucium constellatum ; sequence of events following the excision of the 
distal one-third of the tail; living specimen. Immediately after the beginning of tail resorption 
the tail was excised and photographed. The epidermis began to shorten by contracting over the 
surface of the underlying NMN-complex. Within two minutes the epidermis shortened into an 
annular ring at the base of the tail. The NMN-complex invariably fails to be resorbed in this 
experiment. The epidermis of the distal segment (not shown) also shortens over the under- 
lying tissues following excision. The arrows indicate the cut surface of the epidermis through 
stages of contraction. 



252 



K1CHARU A. CLONEY 



Min 
in p 
10'^ 
KCN 



2 



10 15 20 25 30 

Total period of tail resorption in minutes 



35 



40 



FIGURE 11. Inhibition of tail resorption by potassium cyanide. The total time required for 
the completion of tail resorption was plotted for larvae treated for various times in a solution 
of 10~ 2 M KCN. Experimental larvae were first observed until the first signs of tail resorption 
could be detected. They were then transferred to the KCN solution for from 1 to 6 minutes. 
They were subsequently washed in sea water. Cyanide treatment slows down the rate of tail 
resorption beyond the controls. The range of normal tail resorption is shown by a line on the 
lower left of the graph. The arrow indicates the average time of tail resorption in 14 normal 
specimens. 



dermis of the tail from the underlying notochord-muscle-nerve cord complex 
(NMN-complex). This results in the formation of a fluid-filled subepidermal 
space. The NMN-complex huckles and folds as it moves into the posterior end 
of the trunk. The epidermis forms a thickened cap over the end of the trunk, 
enclosing the other tail tissue. 

3. When the tail was excised after the beginning of tail resorption, the epi- 
dermis was observed to retract independently of the other tail tissues. 

4. The muscle cells do not manifest any tendency to shorten without the 
epidermis. 

5. Potassium cvunide and sodium azide reversiblv inhibit the onset of meta- 
morphosis and slow down the rate of tail resorption if they are applied after the 
beginning of metamorphosis. 

6. Some histological changes in the tail of .linaronchiin and other species of 
ascidians may be the result of the activity of a proteolytic enzyme. 



ASCIDIAN METAMORPHOSIS 25.S 

LITERATURE CITED 

BERRILL, N. J., 1947. Metamorphosis in ascidians. /. AIor/>h.. 81 : 249-267. 

CLONEY, R. A., 1961a. Observations on the mechanism of tail resorption in ascidians. Amer. 
Zool, 1 : 67-87. 

CLONEY, R. A., 1961b. Changes in the fine structure of striated muscle cells during' meta- 
morphosis in ascidian larvae. Aunt. Rcc., 139: 217-218. 

CONKLIN, E. G., 1931. The development of ccntrifuged eggs of ascidians. /. Exp. Zool., 60: 
1-119. 

GRAVE, CASWELL, 1921. Amaroucium constcllatnin (Verrill) II. The structure and organiza- 
tion of the tadpole. /. Morf>li., 36: 71-91. 

GRAVE, CASWELL, 1935. Metamorphosis of ascidian larvae. Papers from the Tortugas Lab. 
Cam. lust. Il\,sli. Puhl., No. 452: 209-292. 

HOFFMANN-BERLIXG, HARTMUT. 1960. Other mechanisms producing movements. In: Com- 
parative Biochemistry, Marcel Florkin and Howard S. Mason, editors. Vol. 2, pp. 
341-370. Academic Press, New York. 

JACKSON, SYLVIA FITTON, 1958. Some aspects of morphogenesis in ascidians. Biol. Bull.. 115: 
335. 

LYNCH, W. F., 1961. Extrinsic factors influencing metamorphosis in bryozoan and ascidian 
larvae. Anicr. Zool., 1: 59-66. 

LUFT, JOHN, 1961. Improvements in epoxy resin embedding methods. /. Cell Biol., 9: 409-414. 

PEARSE, A. G. E., 1960. Histochemistry. Little, Brown and Co., Boston. 

RICHARDSON, K. C, L. JARETT AND E. H. FINKE, 1960. Embedding in epoxy resins for ultra- 
thin sectioning in electron microscopy. Stain Tech., 35 : 313-323. 

SCOTT, SISTER F. M., 1946. The developmental history of Ainaroccinin cnnstclhitnin. II. Or- 
ganogenesis of the larval action system. Biol. Bull., 91 : 66-87. 

SCOTT, SISTER F. M., 1952. The developmental history of Ainaroccinin constcllatuiu. III. 
Metamorphosis. Biol. Bull. 103: 226-241. 

WAINIO, W. W., AND S. J. COOPERSTEIN, 1956. Some controversial aspects of mammalian cyto- 
chromes. Adv. in Enzyin., 17: 329-392. 

\\ EISS, PAUL, 1928. Experimented Untersuchungen iiber die Metamorphoses der Ascidien II. 
Versuche iiber der Mechanismus der Schwanzinvolution. Biol. Zcntr., 48: 387-407. 



MOLTING AND CYCLIC ACTIVITY IN CHROMATOPHOROTROPINS 
OF THE CENTRAL NERVOUS SYSTEM OF THE BARNACLE, 

BALANUS EBURNEUS 1 

JOHN D. COSTLOW, JR. 
Duke University Marine Laboratory, Beaufort, North Carolina 

Although molting in the acorn barnacles is more frequent than that described 
for most Crustacea, every 2-3 days in the warm-water species, and continues 
throughout the life of the individual barnacle (Costlow and Bookhout, 1953, 1956), 
nothing is known of the mechanisms which control the sequence of molting or the 
duration of the intermolt period (Carlisle and Knowles, 1959). Barnes and Conor 
(1958a) have suggested that the control of molting in the Cirripedia may depend 
on the absence of a molt-inhibiting hormone and thus be more similar to the 
mechanism described for Lysinota seticaudata (Carlisle, 1953a, 1953b, 1953c; 
Carlisle and Dohrn, 1953) than the mechanism described for most decapods which 
is thought to involve the interaction of a molt-accelerating and a molt-inhibiting 
hormone (Passano, 1960, 1961). 

Several attempts have been made to establish a relationship between cyclic ac- 
tivity of barnacle shell and body tissues and the regular sequence of molting 
periods. While cyclic activity has been described from a few tissues in barnacles 
(Thomas, 1944; Costlow, 1956), the activity has not been shown to be directly 
associated with either ecdysis or its control. 

The present study has had two major objectives: (1) to determine if the Uca 
black-pigment-dispersing substance of the central nervous system of acorn barnacles 
(Sandeen and Costlow, 1961) is cyclic in its activity; and (2) if activity of this 
chromatophorotropin is cyclic, to determine whether the changes in activity are 
associated with the molting cycle of the adult barnacle. 

METHODS 

Adults of Balanus eburnens were collected from pilings in the area of the Duke 
University Marine Laboratory and placed in the running sea water system. When 
it had been determined that the animal had not been damaged in the course of 
removal from the pilings, the barnacles were segregated into plastic compart- 
mented boxes, placed at a slight angle under the direct flow of the sea water outlet, 
and fed Artcinia nauplii. The barnacles were checked at regular half-hour inter- 
vals during the day and if a molt were found, the time was recorded and the animal 
removed to another container in running sea water. 

At the intervals of time after molting shown in Table I the barnacle body was 
removed from the shell, dried on filter paper, and weighed to the nearest 0.1 mg. 
on a Roller-Smith balance. The central nervous system was then dissected intact 

1 These studies were aided by a contract, 104-194, between the Office of Naval Research 
and Duke University. 

254 



ENDOCRINE CYCLES IN BARNACLE MOLT 

TAIU.K I 

.\ninhcr oj nilii/l lidldtuis eburneus from whali central nervous systems were e.\ir/u Inl <it 
following molting and the dilutions of extract (CNS/ml) used 



255 





Intervals 




Hours after molting 


I >ilution 





12 


24 


48 


72 


96 


1:1 


12 


12 


12 


11 


8 


9 


1:5 


19 





20 


10 


9 


9 


1:10 


9 





9 












from the body and thoroughly triturated in a glass dish containing as little water 
as possible. The ground material was suspended in the desired amount of filtered 
sea water, boiled, and centrifuged. 

Fiddler crabs, Uca pugilator, which had been destalked 12 to 24 hours pre- 
viously, were then injected with the barnacle central nervous system extract. The 
standard dose was 0.05 cc., injected into the ventral hemocoel at the base of the 
fifth walking leg with a tuberculin syringe and a 26-gauge needle. Three different 
concentrations of extract were used : one central nervous system per one ml. of 
sea water, one per five ml. and one per ten ml. For control injections, homoge- 
nized opercular muscle from the barnacles was used and 5 fiddler crabs were in- 
jected daily with Uca eyestalk extract, one pair per 5 ml., to determine the 
variability of the bioassay animals as well as the technique. 

Five eyestalkless fiddler crabs were injected with the extract from each central 
nervous system and the chromatophores of each animal staged at intervals of 15, 30, 
45, 60, 90. 120, and 180 minutes following injection. The chromatophore scale 
of Hogben and Slome (1931) was used to determine the Uca chromatophore re- 
sponse to the barnacle central nervous system extracts. From the average 
chromatophore values for each reading the total net activity was determined for 
the entire ISO-minute period (Sandeen and Costlow, 1961). All experiments 
were begun between 1 and 2 PM and terminated before 5 PM to avoid possible 
fluctuations in the crab chromatophores due to diurnal rhythms. 

RESULTS 

Figure 1 gives the average total activity values for extracts of B. cburncits 
central nervous systems removed at regular intervals following ecdysis and as- 
sayed on eyestalkless fiddler crabs. At a concentration of 1 : 1 the central nervous 
system had a low activity immediately following molting, followed by an increase 
in activity at 24 hours after molting. Statistical treatment of the data from these 
experiments indicates that variability in activity of the extracts within the sample 
is as high as that between the samples, and that the fluctuations in central nervous 
system activity at different periods of time after ecdysis are not significant at a 
concentration of 1:1. 

Injections of the barnacle central nervous system extract at a dilution of 1:5 
showed similar changes in the activity of the chromatophorotropins in relation to 
the molting cycle : a low point of activity immediately following molting, a peak 



256 



JOHN D. COSTLOW, JR. 



2 I 
20 
I 9 
18 

i n 



" 



CTi2 

h- 

X 

LJ ' ' 

to 



U 



I 
9 
8 
1 
6 





12 24 36 48 60 12 

HOURS AFTER MOLTING 



96 



FIGURE 1. Average activity values of Balanus cburncus central nervous systems extracted at 
intervals following ecdysis and analyzed on eyestalkless fiddler crabs, Uca pngilator: solid 
circles, 1: 1 dilution; open circles, 1:5 dilution ; and solid diamonds, 1 : 10 dilution. 

of activity at 24 hours after molting, followed by a decline in activity to the initial 
low point, and a slight increase in activity at 72 and 96 hours after molting. These 
data, when subjected to an analysis of variance, are highly significant at the 95 /o 
level. 

Dilutions of the barnacle central nervous system extract of 1:10 also showed 
an activity which was low at the time of molting. This was followed by an in- 
crease at 24 hours after molting but insufficient data were available for additional 
time intervals to make further comparisons. Table II gives the total activity 
values for all dilutions of extracts used in the experiment. 

Injection of barnacle opercular muscle extracts did not produce any reaction in 
the eyestalkless fiddler crab chromatophores. Day-to-day fluctuations in the 
chromatophore activity of the Uca controls (Uca injected with extracts of Uca 
eyestalks) were slight and never of sufficient magnitude to affect significantly the 
activity values obtained for the barnacle extracts. 

Groups of barnacles from which central nervous systems were assayed repre- 
sented a range of body weights from 24 mg. to 160 mg. Although this does not 
exceed a factor of 10 it was considered desirable to be sure that any consistent 



HXDOCRIXK CYCLES IN BARNACLE MOLT 



257 



TAIU.K 1 1 

I'olul uilivity values of central nervous systems of /!n///n/is clninn'iis 
extracted nl intervals within tnir molting period 





Hours after molting 







12 


24 


48 


72 


96 


I : 1 Dilution 


9.2 


22.4 


24.8 


24.4 


18.8 


20.6 




21.8 


22.6 


23.8 


23.6 


12.4 


20.8 




11.2 


19.6 


24.7 


22.4 


19.4 


21.8 




19.8 


17.8 


9.6 


24.0 


22.4 


15.0 




21.6 


18.4 


22.8 


20.4 


15.2 


22.0 




20.2 


20.2 


18.6 


12.8 


11.2 


9.0 




19.4 


18.4 


20.8 


19.8 


16.4 


18.0 




11.2 


24.0 


20.0 


11.6 


20.4 


21.6 




20.6 


21.4 


21.8 


18.6 




19.6 




11.4 


20.4 


17.0 


23.4 








21.3 


14.6 


24.8 


20.2 








19.6 


22.0 


24.4 








Average 


17.3 


20.2 


21.1 


20.1 


17.0 


18.7 


1 :5 Dilution 


5.8 




16.0 


5.4 


11.0 


13.4 




14.6 




15.2 


13.0 


9.6 


13.0 




8.2 




14.0 


15.4 


11.2 


13.4 




7.2 




11.0 


10.4 


9.8 


12.6 




9.0 




12.4 


10.2 


8.6 


10.0 




9.6 




14.4 


9.8 


14.8 


12.6 




7.4 




14.2 


13.2 


10.6 


12.2 




11.3 




17.6 


13.4 


11.8 


12.4 




8.0 




11.4 


7.6 


16.0 


11.6 




14.8 




14.3 


14.0 








13.5 




18.2 










12.3 




6.0 










18.4 




10.8 










9.6 




18.2 










12.8 




17.8 










16.4 




18.8 










13.2 




17.6 










11.4 




15.6 










3.6 




16.4 








Average 


11.0 




15.5 


11.2 


11.5 


12.4 


1:10 Dilution 


6.8 




11.2 










10.6 




2.4 










9.0 




3.8 










3.8 




10.2 










6.4 




9.6 










9.4 




14.8 










6.6 




11.0 










2.3 




8.6 










0.0 




12.8 








Average 


6.1 




9.4 









JOHN D. COSTLOW, JR. 

change in the level of central nervous system activity was not solelv a function of 
body sixe or weight. Accordingly, the data tor central nervous svsteni activity 
and the corresponding barnacle body weights were analyzed. The statistical treat- 
ment of these data indicates that body weight and activity of the chromatophoro- 
tropins within the central nervous system are independent. 

* 

DISCUSSION 

Although ecdysis in the. barnacles, or Cirripedia, is analogous to the process 
described in considerable detail for the Decapoda, there is little evidence to date 
that the control of molting or the mechanisms of control in these two groups of 
Crustacea are similar. Attempts have been made to compare molting in the 
Cirripedia and in the Brachyura, either by studying possible sites of endocrine 
activity or storage organs or through investigations of physiological responses 
which may be similar. Most of our knowledge of the control of molting or the 
mechanisms of control in barnacles, however, is restricted to that gleaned from 
negative results. 

Costlow and Bookhout (1958), studying the relationship between metabolic 
rate and the molting cycle in Balanus auiphitritc. found that while variations in 
oxygen consumption did occur within any one molting period, there was no indi- 
cation that an increase in respiratory rate occurred prior to molting, comparable 
to the two-fold increase described for some decapods (Roberts, 1957a, 1957b). 
Crisp and Patel (1960) described an endogenously regulated anecdysis in Balanus 
balanoides but Barnes (1962), working with the same species of barnacle, has 
suggested that the molting rhythm of this species is determined to a greater extent 
by environmental factors than endogenous factors. While temperature, diet, 
abundance of food, and reproductive state have been shown to affect the fre- 
quency of molting, at least one environmental factor which affects molting in 
some decapods, photoperiod, does not influence the molting frequency of barnacles 
(Costlow and Bookhout, 1956). 

Evidence for a series of cyclic cellular changes within each molting period, com- 
parable to those described for decapods by Drach (1939), has been scant. Thomas 
(1944) described the cyclic activity of the sublingual and subesophageal glands 
in Balanus perforatits and, while the cycles were concurrent, there was no indica- 
tion that the activity of the glands was even indirectly involved in the control of 
molting. Costlow (1956) reported a secretory cycle in the shell-forming tissues 
in Balanus improvisus but cellular activity, as well as subsequent shell growth, was 
not associated with the 2-3-day molting frequency. Barnes and Conor (1958a, 
1958b) found neurosecretory cells in the central nervous system of several species 
of barnacles and gave a detailed description of the types of cells in Pollicipcs 
polymerus, but did not associate the cyclic activity of these cells with the molting 
cycle. They also studied sections of whole bodies of /'. poly in cms but were unable 
to locate any storage organs for neurosecretory material and considered the pos- 
sibility that discharge of such products may occur directly from the cells into the 
perineural blood sinus. 

Sandeen and Costlow (1961) described one basic similarity which does exist 
between the Cirripedia and the Decapoda. They found two chromatophorotropins 
in the central nervous system of Balanus cbumciis, Chclonobia pahila, and Lcpas 



ENDOCRINE CYCLES IN BARNACLE MOLT 

sp. which were similar in activity to the I'ca black-dispersing substance and the 
Palacinonetes red-pigment dispersing material. The function ot these activators, 
especiallv in the absence of either chromatophores or evcstalks. or their possible' 
role in molting of barnacles was not investigated, however. 

In the present study it has been established that activity of the one chromato- 
phorotropin within the central nervous system of B. cbnnieus, the Uca black-dis- 
persing substance, is cyclic and that the cycle does conform to that demonstrated 
for the two-three-day molting cycle. The greatest changes in level of activity are 
those at the time of molting and at 48 hours after molting, when the activity is 
lowest, and at 24 hours after molting when the activity is highest (Fig. 1). The 
slight increases in activity at 72 and 96 hours after molting are not statistically 
significant but do deserve some comment. Normally, at the temperatures under 
which the experimental barnacles were maintained (23-26 C.), the frequency of 
molting is every 2-3 days. Thus, at 72 hours after the initial molt the barnacle 
would normally have begun the next molt and the activity of the chromato- 
phorotropin would be low (Fig. 1). At 96 hours after the first molt the animal 
would normally have completed the next molt and the activity might be expected to 
equal the peak which corresponds to that shown for the 24-hour interval following 
ecdysis. In animals which do not initiate the mole or complete it within the 
regular 2-3-days frequency the activity of the central nervous system extracts 
never attained the peak described for the 24-hour interval following molting. 

While there is no evidence that the chromatophorotropins of the central nervous 
system of barnacles are directly associated with the control of molting, the present 
study does suggest several possible functions and relationships. If we accept the 
suggestion of Barnes and Conor (1958b) that in the absence of a storage site the 
neurosecretory products of the central nervous system are released directly into 
the hemolymph. it is possible to carry the concept of hormonal control of molting 
in the barnacles one step further. In the absence of a storage site the neuro- 
secretory products are accumulated in the central nervous system itself. Secretory 
activity of the cells would increase following molting and reach a peak at 24 hours 
following ecdysis, the point of highest activity and presumably the highest con- 
centration. This would he followed by a gradual release of the material into the 
hemolymph during the next 24-hour period until the initial low level was again 
reached in the central nervous system. If these products were to include a molt- 
accelerating substance, the release of material into the hemolymph between 24 and 
48 hours following molting would serve to stimulate those additional physiological 
processes which are prerequisite to the actual process of molting. Following 
ecdysis the cycle would be repeated, continuing the 2-3-day frequency of molting. 

In barnacles which have not molted for a second time within 96 hours following 
the previous molt, the absence of an increase in activity of the chromatophorotropins 
of the central nervous system, comparable to that found at 24 hours after molting, 
can be attributed to at least one of two things : ( 1 ) secretory activity within the 
central nervous system may have been insufficient to accumulate products equal 
to the titer found during the previous interval and thus molting is not initiated, or 
(2) the general physiological level of the animal, or the general metabolic state, 
had been reduced by environmental factors in the laboratory, and all functions, 
including molting, were retarded. Because the central nervous system extract of 



200 JOHN D. COSTLOYV, JR. 

anv one barnacle can only he analysed for one time interval within any one- molt, 
it is not possible to follow the entire cvcle of activity of the chromatophorotropins 
within anv one harnacle. 

Further studies obviously arc needed to delineate the various components of the 
central nervous system extracts and determine, either by injection or implantation, 
if any of the various fractions can affect molting in the Cirripedia or the Brachyura. 

SUMMARY AND CONCLUSIONS 

1. The central nervous system of the barnacle, Balaints clninicits, was removed 
at known intervals following ecdysis, extracted in sea water, and assayed by in- 
jecting into eyestalkless Uca pug Hat or to determine if the barnacle chromato- 
phorotropins exhibited a cyclic activity associated with molting. 

2. A cyclic pattern of activity was observed within one intermolt period. The 
changes in activity of the Uca black-pigment-dispersing substance, highly sig- 
nificant at the 95% level, were the low concentrations immediately following molt- 
ing and 48 hours after ecdysis and the high concentration which occurred at 24 
hours after molting. In barnacles which did not molt again within the usual 72-96- 
hour period following the first molt, the level of activity of the central nervous 
system extracts remained low. Body weight and the activity of the central 
nervous system extracts were found to be independent. 

3. The hypothesis is presented that in the absence of a storage organ com- 
parable to the sinus gland of Brachyura, neurosecretory products originating from 
the central nervous system are released directly into the blood. The release of 
these products following the period of greatest concentration, 24 hours after 
molting, stimulates the physiological processes and cellular changes which culminate 
in ecdysis. 

LITERATURE CITED 

BARNES, H., 1962. So-called anecdysis in Balanns balanoides and the effect of breeding upon 
the growth of the calcareous shell of some common barnacles. Lininol. Occtnwt/r., 1 : 
462-473. 

BARNES, H., AND J. J. CONOR, 1958a. Neurosecretory cells in the Cirripede, Pollicipcs pol\- 
incnis J. B. Sowerby. /. Mar. Res., 17 : 81-102. 

BARNES, H., AND J. J. CONOR, 1958b. Neurosecretory cells in some cirripedes. Nature, 181 : 
194. 

CARLISLE. D. B., 1953a. Studies on Lysinata seticaudata Risso (Crustacea Decapoda). III. 
On the activity of the moult-accelerating principle when administered by the oral route. 
Pubbl Stas. Zool Napoli, 24 : 279-285. 

CARLISLE, D. B., 1953b. Studies on Lysinata seticaudata Risso (Crustacea Decapoda). IV. 
On the site of origin of the moult-accelerating principle experimental evidence. 
Pubbl. Stas. Zool. Napoli, 24 : 285-292. 

CARLISLE, D. B., 1953c. Studies on Lysinata seticaudata Risso (Crustacea Decapoda). VI. 
Notes on the structure of the neurosecretory system of the eyestalk. Pubbl. Staz. 
Zool., Napoli, 24: 435-447. 

CARLISLE, D. B., AND P. F. R. DOHRN, 1953. Studies on Lysinata seticaudata Risso (Cru- 
stacea, Decapoda). II. Experimental evidence for a growth and moult-accelerating 
factor obtainable from eyestalks. Pubbl. Stas. Zool, Napoli, 24: 69-83. 

CARLISLE, D. B., AND F. KNOWLES, 1959. Endocrine Control in Crustaceans. Cambridge Uni- 
versity Press. 

COSTLOW, J. D., JR., 1956. Shell development in Balamis improvisns Darwin. J. Morph., 99 : 
359-415. 



ENDOCRINE CYCLES IN BARNACLE MOLT 261 

COSTLOW, J. D., JR., AMI C. G. BOOK 11 OUT, 1953. Moulting and growth in Btihiuux iiufruristix. 

Biol. Hull., 105 : 420-433. 
COSTLOW, J. D., JR., AMI C. G. UooKiiorT, 1956. Molting and shell Drouth in Btilnniis tunplii- 

tritc uircus. Biol. Bull., 110: 107-116. 
COSTLOW, J. D., JR., AND C. G. BOOKHOL'T, 1958. Molting and respiration in Buhmux uniplii- 

tritc var. dcitticuluta Broch. Physiol. Zool., 31: 272-280. 
CRISP, D. J., AND B. S. PATEL, 1960. The moulting cycle in Bulauus Inildimidt-s L. Biol. Bull., 

118: 31-47. 
DRACH, P., 1939. Mue et cycle d'intermue chez les Crustaces decapodes. .-Inn. Just. Occanni/r.. 

Monaco, 19: 103-391. 

HOGBEN, L. T., AND D. SLOME, 1931. The pigmentary effector system. VI. The dual charac- 
ter of endocrine coordination in amphibian colour change. Proc. R\. Soc. L<nuln. 

Scr. B, 108: 10-53. 
PASSANO, L. M., I960. Molting and its control. In: The Physiology of Crustacea, vol. 1: pp. 

473-536. Academic Press, New York and London. 

PASSANO, L. M., 1961. The regulation of crustacean metamorphosis. Amcr. Zool., 1: 89-95. 
ROBERTS, J., 1957a. Thermal acclimation of metabolism in the crab, Pachygrapsus crnssipcs 

Randall. I. The influence of body size, starvation, and molting. Ph\siol. Zool., 30: 

232-242. 
ROBERTS, J., 1957b. Thermal acclimation of metabolism in the crab Pachygrapsus crassipcs 

Randall. II. Mechanisms and the influence of season and latitude. Physiol. Zool., 30: 

243-255. 
SANDEEN, M. L, AND J. D. COSTLOW, JR.. 1961. The presence of decapod-pigment activating 

substances in the central nervous system of representative Cirripedia. Biol. Bull.. 120: 

192-205. 
THOMAS, H. S., 1944. Tegumental glands in Cirripedia Thoracica. Quart. J. Micr. Set., 84: 

257-282. 



PROPRIOCEPTION IN THE LEGS OF PHALANGIDS 

ARLAN L. EDGAR 
Department of Biology, Ahna College, Alma, Michigan 

The presence of thin areas in the exoskeleton of certain arachnids and other 
arthropods has been known for about 80 years. The thin areas, made up of epi- 
cuticle, are usually bordered by a thickened lamella and occur in two general 
shapes slit-like and circular. Isolated, slit-like thin areas were noted by Bertkau 
(1878), while groups of more or less parallel slits resembled the ancient stringed 
instrument, the lyre, and prompted the term, "lyriform organs" (Gaubert, 1890). 
The circular ones were first described by Berlese (1909) and called "campaniform 
sensillae." Typically, the dendritic process of a bipolar neuron is attached to the 
epicuticle or to the bordering lamella. Snodgrass (1935) and Kaston (1935) 
have summarized the histology of the former condition in insects. In 1938 Pringle 
presented evidence from the cockroach that indicated stress reception as the real 
function. 

Recently, the attachment of the neurite to the bordering lamella has been de- 
scribed with electron micrographs by Salpeter and Walcott (1960), and shown 
to function as a vibration receptor in spiders. The neurite is stimulated by stretch- 
ing in the region of attachment to the epicuticle. Most commonly this is done by 
compression pressure on the bordering lamella. This results in exaggerating the 
convexity of the epicuticle. 

These organs exhibit considerable morphological variation and have been re- 
ported in insects, spiders, scorpions, mites, ticks and phalangids (Gaubert, 1892; 
Hansen, 1893; Hansen and Sorensen, 1904). The present paper illustrates cam- 
paniform and slit organs (isolated, grouped, and lyriform) of phalangids and re- 
ports evidence suggesting the function of at least the campaniform organs to be 
proprioceptive. 

PROCEDURE AND RESULTS 

Legs of phalangids possess several shapes and combinations of sensillar organs. 
The location and number on each of the four legs for six species are indicated in 
Table I. Considerable similarity exists among all of the species shown except 
Caddo agilis. In addition to the sensillar arrangement, this form is distinguished 
from the others in Table I by size, habitat and morphology. 

Data for all the leg segments are included except for the coxa. On this seg- 
ment, typically a single slit sensillium occurs on the distal margin. At the distal 
margin of the trochanter there occurs, except in Caddo, a lyriform organ with a 
fairly constant number of slits- seven, for example, in Phalangium opillo (Fig. 1). 
Small clusters of I- and L-shaped slit sensillae occur on the proximal portion of the 
femur (Figs. 1 and 2). The leg autotomy plane is located between the trochanter 
and femur. Presumably, the rich sensillar supply on these segments functions to 
indicate to the animal mechanical stress upon this plane. Large, single slits are 

262 



PKOPklOCHI'TlOX IN I'll. M. \\(, IDS 



263 



TABLE I 

Location and number of proprioceptor organs on the legs of certain phalangids 



Leg segment 


Caddo agilis 
Leg number 


Opilio parii'liiiifi 
Leg number 


PhaliiiiRiHm npilio 
Leg number 


1 


2 


3 


4 


1 


2 


3 


4 


1 


2 


3 


4 


Trochanter 














4 


2 


4 


5 


7 


7 


7 


7 


Femur P* 


9 


9 


9 


5 


18 


16 


22 


21 


29 


30 


30 


29 


-S* 


1 


1 


1 


1 


2 


4 


2 


2 


4 


5 


4 


4 


-D* 


1 


1 


1 


1 


1 


1 


1 


1 


1 


1 


1 


1 


Tibia P* 


2 


p 


2 


2 


6 


3 


6 


6 


9 


8 


8 


9 


Metatar. P* 










7 


5 


7 


7 


9 


9 


8 


9 


Tarsus 1st* 
Mid.* 












1 


1 
6 


6 


1(2) 
7-12 


1(2) 
11-14 


1 
9-10 


1(2) 
7-H 


Leg segment 


Li'iobniinin (alcar 
Leg number 


Leiobunum longipes 
Leg number 


Leiobunum pallium 
Leg number 


1 


2 


3 


4 


1 


2 


3 


4 


1 


2 


3 


4 


Trochanter 


5 


5 


5 


5 


6 


7 


7 


5 


5 


5 


4 


6 


Femur P* 


21 


24 


23 


24 


30 


31 


30 


27 


23 


27 


22 


29 


-S* 


4 


5 


4 


5 


4 


5 


4 


4 


3 


5 


4 


4 


-D* 


1 


1 


1 


1 


1 


1 


1 


] 


1 


1 


1 


1 


Tibia P* 


7 


5 


5 


7 


7 


7 


7 


6 


7 


7 


7 


9 


Metatar. P* 


8 


6 


7 


8 


8 


8 


8 


8 


8 


10 


8 


10 


Tarsus 1st* 


1 


1 


1 


1 


1 





1 


1-2 


1 


1 


1 


1 


Mid.* 


7 


7 


8 


6-8 


8-11 


13-16 


6-10 


10-11 


4 


9 


6 


7-8 



* Key to symbols : 

P organs located on proximal portion of leg segment. 
S shank of leg segment. 
D distal portion of leg segment. 

1st number of organs found on the most proximal article of tarsus. 

Mid. the number of the article on which is found a campaniform organ. Extremes (for 
example, 7-12) indicate that organs have been found on articles within these limits. 



more or less evenly spaced along the shaft of the femur. The largest slit organ 
on the leg occurs on the distal femur where it articulates with the patella (Fig. 3). 
The ventral, proximal portion of the tibia has a cluster of campaniform sensillae. 
Usually, three slit organs and three to seven campaniform sensillae occur on the 
dorsal surface of the metatarsus near the articulation with the tibia. Occasionally 
one or two of these latter organs are separated and distal from the cluster. Typi- 
cally, two solitary campaniform organs are found on the tarsus ; one is on the 
proximal-most article and the other on an article in the middle one-third of this 
segment (Fig. 4). Observations of the organs tabulated in Table I were made 
from exuviae of the final molt. 

Although Savory (1962) and others have indicated that the second pair of 
legs possess special senses, the arrangement and kinds of sensillar organs listed 
in Table I do not show any consistent disparity with the other three pairs. Legs 
from all four pairs were used in the experimentation ; however, no differences be- 



264 



ARLAX L. EDGAR 




FIGURE 1. Lyriform organ at distal margin of trochanter (arrow). The autotomy plane 
which separates trochanter from femur is shown under the other end of the arrow. A cluster 
of I- and L-shaped slit organs may he seen on the proximal portion of the femur fright) . 

FIGURE 2. Several slit organs clustered at base of femur (see Figure 1). The expanded 
area in the middle of the slit probably receives the attachment of a bipolar neuron. 

FIGURE 3. Large, thick-bordered slit organ on femur (arrow) where femur articulates 
with patella. 

FIGURE 4. Campaniform sensillium on twelfth article of the tarsus ; the structure measures 
approximately 20 by 15 microns. Figures 1-4 are photographs made from molt cases of 
Phalangiiun opilio L. 



tween legs were observed in the records obtained. Oscilloscope traces shown in 
Figures 5-7 were obtained from the left leg I of Plialanc/imu opilio. 

Legs from health}- adults were removed at the autotomy plane between the 
trochanter and femur and mounted upon a Plasticene support. Platinum wire 
electrodes inserted into the open, proximal end and the patellar end of the femur 
were connected via a Grass Type P. 4 preamplifier to a Tektronix 531 oscilloscope. 

By adequate support of all sections of the leg except the test area, stress could 
be placed on specific sensillae. Stress was produced by hyperextending or hyper- 
flexing the normal angle of movement at an articulation nearest the sense organ. 
This was done by manually-operated glass and steel probes. 

Although pressure was applied on all joints of the leg, action potentials were 
consistently obtained only upon hyperextension of the tibiometatarsal joint (Fig. 5) 
and hyperflexion of the tarsometatarsal joint (Fig. 6) and tarsal articles. Adapta- 
tion was rapid and complete. 

To localize the site of origin of the nervous activity shown in Figures 5 and 6, 



PROPRIOCEPTION IN PHALANGIDS 



265 




FIGURES 5 and 6. Oscilloscope records from leg nerve at the femur. Figure 5. Forced 
extension of the tibiometatarsal joint of P. opilio L. Figure 6. Forced flexion of a tarsometa- 
tarsal joint of the same species. Arrows indicate time at which stress was applied. 

FIGURE 7. Three oscilloscope records of extension of tarsometatarsal joint. The typical 
pattern of this trace may be seen in Figure 6 near the right-hand side. Arrows indicate the 
moment of flexing ; the time bar shows one-second interval on Figures 5, 6 and 7. 

articles of the tarsus and metatarsus were removed, a few at a time, by cutting. 
Impulses continued to be elicited, upon stress, in a normal manner until the im- 
mediate region of a sensillium was removed ; then the record ceased. Local 
destruction attempts were unsuccessful. 

A series of differently-shaped impulse discharges usually appeared on the 
record about 0.4 second after adaptation (Fig. 6). It was interesting to note that 
a wave form which had a similar shape appeared upon moving the leg segment in 
the direction opposite to that producing the stress discharges and without ex- 
ceeding the normal limits of movement. Figure 7 indicates three such records 
from extension of the tarsometatarsal. Just where this second discharge originated 
is unknown, possibly in a different sense organ located nearby, or perhaps two 
neurons are associated with the same sensillium to transmit "action" and "reaction" 
stimuli. However, morphological descriptions known to the writer have shown 
only one nerve cell involved for each organ. 

Attempts to induce distinct responses from the grouped slits (Figs. 1 and 2 ) 
generally failed. However, surface pressure from careful scraping of the area 
with glass and steel probes resulted in a jumble of electrical activity. \Yhen nearby 
spines were intentionally prodded, no response was obtained. This was sufficient 
to suggest that impulses were coming from the slit organs. These manipulations 
were observed plainly under suitable magnification. 

The slit sensillae of the femur, with one exception, have yielded no clear re- 
sponse to stimulation. On this one occasion it is believed that insertion of the 
electrode into the femoropatellar junction caused pressure to activate the large 
single slit sensillium at that joint (Fig. 3). Very regular, large, fast impulses con- 
tinued for many minutes. They presumably did not originate from leg segments 
other than the femur since manipulation of these segments caused no alteration in 
the frequency or amplitude of these discharges. Subsequent repositioning of the 
femoropatellar electrode abolished the response. It is possible that one of the other 
sensillae located on the shank of the femur was the organ responding. This is 
unlikely, however, in that lateral movements of the shank, designed to place stress 
upon these organs, produced no response. 



266 ARLAN L. EDGAR 

Response to parameters other than mechanoreception was sought. Sound 
waves from a tuning fork ( 100 cps) and loud noises resulted in no detectable re- 
sponse by the intact, functioning sensillae. Addition of xylene to the tarsal and 
metatarsal sensillae caused no observable response. These sense organs functioned 
typically before and after application of the xylene. Pringle (1955) used the 
application of liquid xylene on slit sensillae as a criterion of possible chemoreceptive 
sensitivity. 

DISCUSSION 

The shape of phalangids is exaggerated as compared to that of most arthropods, 
and indeed other forms, in that the legs typically are extremely long and of small 
diameter compared to the small, oval body. Being without antennae the animal 
first encounters its environment predominately with its legs. For this reason, one 
might expect to find a variety of sensory organs here. 

The legs are directed radially and upward from the animal in an arch so that 
the body is suspended in the middle third of the distance from the surface touched 
by the tarsi to the highest part of the arch of the leg. The body is therefore ex- 
posed only directly above and below. All joints of the legs operate essentially in 
one plane with respect to the body (dorso-ventral) except the trochanter-femur 
which, in combination with the coxa-trochanter, is capable of movement in a 
variety of dorso-ventral and anterio-posterior directions. It is at the femoro- 
trochanter junction that the leg may be autotomized. In nature, phalangids fre- 
quently are encountered with one or several legs missing. Legs grasped by 
predators or trapped during the molting process are readily shed in order to escape. 

Sensillar supply is richest near this autotomy plane and diminishes to where the 
only type present on the distal half of the leg is the campaniform sensillium. These 
few isolated organs register stretch and compression pressure in the region of 
joints, i.e., between leg segments or between the hinge-like series of tarsal articles. 
Presumably, stress in other than the plane of the joint movement would be de- 
tected mainly in the region of the autotomy plane, where groups of slit sensillae are 
oriented so as to respond to deformation from any direction (Figs. 1 and 2). The 
campaniform organs detect shock stress coming from ventral ( about the tibiometa- 
tarsal joint) and dorsal or ventral directions (about the metatarsotarsal joint and 
adjoining tarsal articles). These organs apparently do not function to indicate, to 
the animal, position of the appendage while at rest or from contacts with the rela- 
tively stationary aspects of the environment (forest litter, tree trunks, etc.) since 
action potentials are picked up only when the normal limits of the joint are 
exceeded. Hence, they function as well-distributed alarm systems which "go 
off" when the safety of the limb and probably the animal are threatened. 

SUMMARY 

1. In the legs of phalangids thin areas in the exoskeleton occur in four basic 
arrangements: (a) solitary slit sensillae. (b) I- and L-shaped slits in clusters, (c) 
lyrifonn organs, and (d) circular or campaniform sensillae. 

1. Evidence was obtained to indicate that at least the campaniform sensillae 
on the phalangid tarsus and metatarsus function as proprioceptors. Action poten- 



PROPRIOCEPT10N IN PHALANC1DS 267 

tials, from grouped and isolated slit organs on the tibia and femur, were elicited 
but were less distinct as to electrical characteristics and function. 

3. No response was seen upon application of sound waves and xylcne to the 
cninpunitt >rm sensillae. 

LITERATURE CITED 

BERLESE, A., 1909. Gli Insetti, 1. Soc. Edit. Libraria, Milan. 

BERTKAU, P., 1878. Versuch einer naturlichen Anordnung der Spinnen nebst Bemerkungen zu 
einzelnen Gattungen. Arch. Naturgesch., 44: 354-510. 

GAUBERT, P., 1890. Note sur les organes lyriformes des Arachnides. Bull. Soc. Philoin. 
Paris, Scr. 8, 2 : 47. 

GAUBERT, P., 1892. Rechercbes sur les organes des sens et sur le systeme tegumentaire, glan- 
dulaire et musculaire des appendices des Arachnides. Ann. Sci. Nat., 13 : 31-185. 

HANSEN, H. J., 1893. Organs and characters in different orders of arachnids. Entomol. 
Mcdd. ndg af Ent. For. i Kjobcnhavn, 4 : 137-251. 

HANSEN, H. J., AND W. SORENSEN, 1904. On Two Orders of Arachnida. Cambridge Uni- 
versity Press. 

KASTON, B. J., 1935. The slit sense organs of spiders. /. Morph., 58 : 189-207. 

PRINGLE, J. W. S., 1938. Proprioception in insects, I. A new type of mechanical receptor 
from the palps of the cockroach. /. Exp. Biol., 15 : 101-113. 

PRINGLE, J. W. S., 1955. The function of the lyriform organs of arachnids. /. E.vp. Biol., 32: 
270-278. 

SALPETER, M. M., AND CHARLES WALCOTT, 1960. An electron microscopical study of a vibra- 
tion receptor in the spider. E.rp. Neurology, 2 : 232-250. 

SAVORY, T. H., 1962. Daddy Longlegs. 5ft. Amer., 207 : 119-128. 

SNODGRASS, R. E., 1935. Principles of Insect Morphology. McGraw-Hill Book Co., Inc.. 
New York and London. 667 pp. 



OBSERVATION ON THE ECOLOGY AND REPRODUCTION OF 
FREE-LIVING CONCHOCELIS OF PORPHYRA TENERA 

HIDEO IWASAKI AND CHIKAYOSHI MATSUDAIRA 

Department of Fisheries, Faculty of Agriculture, Tohokit University, Sciidai, Japan 

In a previous paper (Iwasaki, 1961) it was shown that free-living Conchocclis 
colonies produce normal monosporangia and monospores under short-day condi- 
tions (8-11 hours light daily); these monospores germinate into leafy thalli as 
they do in nature. 

The free-living Conchocclis colonies cultured in continuous light produce only 
other Conchocclis colonies. It hecame then interesting to find how new Conchocclis 
colonies are produced. It was found that new colonies could he ohtained from 
small pieces of Conchocclis filaments. However, the Conchocclis colonies grown in 
continuous light produce sporangia which are morphologically different from the 
normal sporangia produced under short-day conditions. Even though free spores 
were not found in continuous light, it was not excluded that new Conchocclis 
colonies could he produced by special spores produced under continuous light. 

The present work was undertaken to study the type of sporangia formed by 
free-living Conchocclis colonies under long-day conditions, and the fate of their 
spores. It was found that spores are indeed formed and that they can originate 
new Conchocclis colonies, revealing, under these conditions, a new part of the life- 
cycle of P. tcnera. 

MATERIALS AND METHODS 

The original material derives from mature leafy thalli collected at Matsukaw r a- 
ura Inlet located near Sendai in 1960. The Conchocclis colonies used for the 
present studies were obtained from carpospores produced by the leafy thallus 
passing through one life-cycle in vitro. This Conchocclis strain was a uni-algal 
culture, but accompanied by bacteria and yeast. Two treatments with ultraviolet 
light for 1-2 minutes, at intervals of one month, eliminated the bacteria but not 
the yeast. Two types of liquid media, the artificial medium ASP12NTA (Pro- 
vasoli's medium), and the enriched sea water medium SWI, were employed for the 
experiments (Table I). The cultures were carried on mainly in screw-cap tubes 
(125 X 20 mm.) with 10 ml. of medium. When necessary, the hanging culture 
technique was employed. 

RESULTS 
I. Types of sporangia produced in long-day conditions 

a) The "inflated spherical" cells 

Small pieces of Conchocclis filament were inoculated in culture media and 
grown under different conditions. The first observation was made on the strain 
cultured at a window in subdued natural light (max. 500 ft.c.) from July to 

268 



CONCHOCELIS OF PORI'HVRA 



269 



TABLE I 

Enriched sea water medium, S\\ I 



Filtered sea water 

KN0 3 

KH 2 PO 4 

Fe-EDTA (1:1 chelation) 

"Tris Buffer"* 

pH 



1000 ml. 
72.2 mg. 
8.8 ing. 

0.5 mg. (as Fe). 
500 mg. 

7.8-8.0 



Tris (hydroxymethyl) amiuo methane (Sigma Company). 



Artificial medium ASP12XTA 



Distilled water 

XaC) 

MgS0 4 -7H 2 O 



KC1 

Ca (as Cl) 
NaN0 3 

K 3 PO 4 

Na 2 glycerophosphate 



100 ml. 

2.8 g. 

0.7 g. 

0.4 g. 

0.07 g 

40 mg. 

10 mg. 

1 mg. 

1 mg. 



Xa 2 SiO 3 -9H,O 

B 12 

Biotin 

Thiamine 

P II metals* 

S II metals** 

"Tris" buffer 

Xitrilotriacetic acid 

pH 



15 mg. 

0.02 M g. 

0.1 Mg. 
10 Mg- 

1 ml. 
1 ml. 
0.1 g. 
10 mg. 
7.8-8.0 



* One ml. of P II metals contains: EDTA, 1 mg. ; Fe (as Cl), 0.01 mg. ; B (as H 3 BO 3 ), 0.2 mg. ; 
Mn (as Cl), 0.04 mg. ; Zn (as Cl), 0.005 mg. ; Co (as Cl), 0.001 mg. 

**One ml. of S II metals contains: Br (as Na), 1.0 mg. ; Sr (as Cl), 0.2 mg. ; Kb (as Cl), 0.02 
mg. ; Li (as Cl), 0.02 mg. ; I (as K), 0.001 mg. ; Mo (as Na), 0.05 mg. 



September at temperatures ranging from 20 to 28 C. Small plants of Conchocells 
formed new lateral branches in a week. Some of the tips of branches swelled to 
form a more or less spherical cell about three weeks after inoculation. These 
cells became more deeply pigmented and gradually reached 11.4-14.0 \i. in diam- 
eter (Plate I, A). Soon the spherical cells detached from the branches and became 
free spores, mostly spherical in shape (9.8-11.7 /j. in diameter, Plate I, B). The 
spores begin to germinate after 3-7 clays from the liberation. At first the spore 
forms a germ tube, then a delicate cross-wall appears between the original spore 
and the tube, and soon lateral branches form. These early filamentous germlings 
were usually quite tortuous. They did not orient to any particular direction in re- 
lation to light. Finally, the germlings grew into luxuriant branched filamentous 
thalli. 

b) The special ''sporangia" 

In the culture grown at 17-19 C. and exposed to 16 hours fluorescent light 
(350 ft.c.), the Concliocelis colonies produced many sporangia (SI) as shown in 
Plate I, D-F. The sporangia are very similar to the sporangia produced under 
continuous light (Iwasaki, 1961, Fig. 4, p. 178). As mentioned in that paper, the 
sporangia produced under long-day conditions seemed to be different morphologi- 
cally from the monosporangia produced under short-day conditions. The sporangia 
cells have thicker walls and the length of cells is usually about half their diameter. 

r) The "strawberry-like" bodies 

In the mass culture under the same conditions, very strange, round strawberry- 
like structures (70-80 /,. in diameter) were found. These structures (Plate I. C) 



270 



H1UKO IWASAK1 AND CHIKAYOSHI MATSUDAIRA 




PLATE I 



cells (A) and liberated conch.o-inonospore which germinate into 
us (B). < 256. C. Round strawberry-like structures formed on the Con- 
chocelis branches. < 160. D, E, F, G. Sporangia (SI) formed in long-day conditions; 1), > %; 
E, X 192; F, X 160; G, liberation of the spores, X 384. 



A, 15. Inflated spherical cell 
Conchocelis thallus (B). < 



COXCHOCELIS OF PORPHYRA 271 

have numerous very slender colorless filaments, less than 1 //. broad, 10-20 /j. in 
length, on the surface. At first they seemed to be some kind of parasitic fungi, hut 
about a month after their appearance, many new filamentous Conchocelis colonies 
were found in the culture. 

II. Fate of the ''spores" produced by the "strawberry-like" bodies 

Suspecting that these germlings might derive from the "strawberry-like'' bodies, 
16 of these structures were cut off with a scalpel from the branches, then cultured 
in hanging drops (ten structures) and in screw-cap tubes (six structures), re- 
spectively, under identical conditions (long day, 350 ft.c. fluorescent light). These 
structures, even after the separation from the branches, grew to the size of about 
140 p. in diameter. 

(7-1) In two cultures in hanging drops, several germ tubes grew directly from 
the strawberry-like structure after 16 days. The germ tubes elongated, formed 
lateral branches, and then grew into Conchocelis colonies by repeated branching. 

a-2) In two other cultures, after three weeks or more, the structures disinte- 
grated, liberating several ameboid, unicellular spore-like bodies. Two of them 
germinated in a way similar to the carpospores produced by leafy thallus and grew 
into free-living Conchocelis colonies. Two kinds of Conchocelis filaments were ob- 
served, one thick (7.6-8.0 p. broad) and the other thin (4.0-5.0 /j. broad). These 
two kinds of young Conchocelis were transferred to new medium, but they be- 
came infected by bacteria and the culture medium became cloudy. The thick- 
filamented Conchocelis grew nicely despite the bacterial infection and the culture 
medium became clear after three weeks. The thin-filamented Conchocelis bleached 
soon after transfer. 

a-3) The other straw T berry-like structures in hanging drops did not germinate 
and became covered by very slender dark brown filaments after a month or more. 

b-l) The six structures inoculated in screw-cap tubes attached themselves im- 
mediately to the side wall of the tubes. After two weeks the internal cells of the 
structure produced several germ tubes which grew into normal Conchocelis (four 
cultures). One of these Conchocelis colonies formed monosporangia after 45 days, 
and soon liberated monospores that gave rise to leafy thalli reaching 23 mm. be- 
fore they became pale and died under the same long-day conditions. Half of these 
monospores were isolated and also germinated into normal leafy thalli under short- 
day conditions (9 hours daily of incandescent light) at 13-15 C. Three other 
Conchocelis cultures formed very poor sporangia which did not liberate any spores. 
either under long-day or under short-day conditions (150-300 ft.c. fluorescent light 
for 270 days). 

b-2) One of the strawberry-like structures disintegrated in two weeks but 
after a month, eight young Conchocelis colonies were found. Unfortunately, the 
observation of this experiment was done through the wall of the test tube, using 
a dissecting microscope, so the germination could not be followed in detail. It is 
probable that these young Conchocelis colonies originated from ameboid-shaped 
spores produced by the strawberry-like structure. These Conchocelis colonies 
formed many monosporangia and liberated a lot of monospores under short-day 
conditions. These monospores also germinated into normal leafy thalli. 



HIDEO IWASAKI AND CHIKAYOSHI MATSUDAIRA 

III. J'tile t>j the "spores" oj lon</-day sporangia (-S7) under different Hi/lit periods 

Half ot the Conchocelis culture, which formed "parasporangia-type" sporangia 
(SI, see section Ih), were moved to short-day conditions ( ( ) hours daily of 200 ft.c. 
incandescent light) at 13-15 C., and the other half were kept at 17-19 C. under 
fluorescent light of 200 ft.c. and a daily photoperiod of 14 hours. In about two 
weeks, the sporangia discharged many spores. The liberation of the spores was a 
little earlier in the short-da v condition. The manner of spore liberation is shown 
in Plate I, G. These spores were taken up with a capillary pipette and inoculated 
in new culture media. The majority of spores did not germinate either under 
short-day or long-day conditions. In general a distinctly higher percentage of 
germination was observed in closely grouped spores than in less closely grouped 
clusters. Many abnormal pale germlings (Plate II, E, F), which soon bleached, 
were observed. 

a) Under long-day conditions the spores gave rise to filaments but soon the 
original spore enlarged and divided, forming a sporangia-like body (Plate II, A). 
Sporangia-like bodies are also formed in the terminal part of filaments or as 
lateral growth on the singly-branched filaments (Plate II, B). After that the 
filaments became Conchocelis colonies and produced well developed luxuriant 
sporangia-like bodies (Plate II, B). 

Some other spores continue to enlarge after liberation ; occasionally they pro- 
duce root-like projections and grow into globular bodies, about 110 p. in diameter 
(Plate II, C, D). Another group of spores enlarged, then grew into irregular- 
shaped "blades" as shown in Plate II, G. 

b) The spores kept under short-day conditions produce short, colorless fila- 
mentous projections (about less than 100 /*). The projections may occasionally 
branch. The original spore enlarged while the filamentous projections elongated 
and became more deeply pigmented, then formed a cross-wall (Plate III, F). At 
this stage, the filamentous part bleached while the original spore continued to di- 
vide, forming a sporangia-like body composed of cells quite short and broad (Plate 
III, G). Xo more elongation and longitudinal division of germlings occurred even 
under these conditions which are suitable for the growth of leafy thallus. Some 
germlings produce robust branches near the original spore body, occasionally on 
the first sporangia-like branch, and grow into the aberrant plantlets as shown in 
Plate III, A-E. "Massive plantlets" (Plate III, I-K, L-O), irregular-shaped 
blades and abnormal germlings, as mentioned above, were also found in the 
culture under short-day condition. 

DISCUSSION 

New Conchocelis colonies can grow from small pieces of filaments (less than 1 
mm. in length). Some of the new colonies, therefore, may derive from small pieces 
of filaments that are cut off naturally, especially in old cultures, and by shaking. 

Inflated spherical cells are often formed at the tip of branches. These inflated 
cells were at first thought to be undeveloped monosporangia, but no leafy thalli 
developed from the spores found in these cultures. It is sure, therefore, that the 
spores derived from inflated cells are quite different functionally from the mono- 
spores that develop into leafy thallus. It is not clear yet whether or not the inflated 
cells are formed on definite branches. This asexual reproduction closely resembles 



CONCHOCELIS OF I'OKPHYRA 



273 






\ 







& 







PLATE II 

Various germlings developed from "spores" originating from sporangia (SI) in 14 hours' 
illumination daily. A, B, Conchocclis-like plants; A, X 160, B, X 72. C, D, globular body, 
X 160. E, F, (Abnormal one?) X 160. G. Blade-like body, X 160. 



that of Rhodochorton. It would seem fitting to call them "Concho-monospores" to 
specify that they originate new Conchocelis colonies and to differentiate them from 
the normal monospores which develop into leafy thalli. 

It is not clear whether the strawberry-like structures are polysporangia or 
cystocarps. The difference observed in the germinative processes derived from iso- 
lated strawberry-like structures seems mainly due to the maturation stage of the 
structure. The strawberry-like structures matured normally on the CoucJiocdis 
branches, liberating two to eight spores which germinated into Conchocelis colonies. 



274 



HIDEO IWASAKI AND CHIKAYOSHI MATSUDAIRA 




B 





H 



G 



^B^ 



" 















PLATE III 

Various germlings developed from "spores" originating from sporangia (SI) in 9 hours' 
illumination daily (except H). A-E, Plantlets developing from the "spores," X 160. E, X 192. 
F-G, Blade-like bodies, X 160. I-O, Massive plantlets, X 160. H, Young buds of leafy thalli 
developed from monospores, X 160. 



CONCHOCELIS OF PORPHYRA 



275 



These Conchocdis colonies formed monosporangia and liberated monospores which 
germinated into normal leafy thalli under short-day conditions. 

One of the Conchocdis colonies derived from straw! >erry-like structures pro- 
duced monosporangia and monospores which gave rise to leafy thalli ( 1-$ mm. in 
length) even in long-day conditions. This behavior of the Conchocdis phase in 
long-day conditions (16 hours daily, fluorescent light) had never been observed 
in our culture in vitro. Is this Conchocdis phase a special one or a mutant? More 
work is needed to solve these questions. 

The "Concho-monospores" and strawberry-like structures have been observed 
at temperatures higher than 18 C. and under long-day conditions. It seems that 
the formation of inflated spherical cells and the strawberry-like structures is af- 
fected by photoperiodism and temperature. 

The sporangia (SI) produced by Conchocdis colonies grown under long-day 
conditions resemble in function the plantlets that Kornmann used to start his cul- 
tures (1960). These sporangia produce spores which germinate into new Con- 
chocdis colonies having well developed sporangia under long clay, though their 
germination rate is considerably lower. 

In the short-day condition, the spores develop into "plantlets" (ours) having a 
few or poor filaments (rhizoids). The "plantlets" are morphologically very similar 
to the sporangia that Condwcdis develops in long day but seem to be functionally 
different. 

Although numerous "plantlets" with rhizoids appear in the culture of P. pscudo- 
hncaris (Plate IV), this type of "plantlet" without rhizoids has not been found in 
P. tenera cultures. The "plantlets" found in our cultures under short-day condi- 






PLATE IV 
"Plantlet" without rhizoids of Porphyni ^scinl,>lin,;iris. A, X 192; B, X 384. 



276 HIDEO IWASAKI AND CHIKAYOSHI MATSUDAIRA 

linns are similar to the "plantlcts" described by Drc-\v (1954) in her culture, of 
/'. umbilicalis. 

The blade-like bodies developed in sbort day ( I 'late 111, F, G) look like young 
buds of leafy thalli but they do not elongate to more than 1 mm. in length and do 
not show longitudinal divisions (Plate III, F, G and H). 

What are the globular and irregular blade-like bodies (Plate II, C, D and G) 
developed in long-day conditions, and what roles do they play in the life-cycled 
And also, what are the blade-like bodies (Plate III, G) and "massive plantlets" 
(Plate III, I-O) in short-day conditions? Are they merely abnormal germinations 
of leafy thalli? The "plantlet" seen in Plate III, O, seems ready to liberate some 
unknown spores. 

The variety of structures created under different light and temperature condi- 
tions shows that P. tcncra has unusual power of adaptation to the environment, and 
more work is needed to solve some of the problems posed by the growth potencies 
of P. ten era. 



This work was supported in part by Contract NR 104-202 of the Office of 
Naval Research, U.S.A., to the Haskins Laboratories; we wish to thank Dr. L. 
Provasoli for his help and useful suggestions. 

SUMMARY 

1. Free-living Conchocelis colonies under long-day conditions produce several 
types of reproductive bodies from which new Conchocelis colonies originate, re- 
vealing a new part of the life-cycle. 

2. Three reproductive structures were observed and described: (a) inflated 
spherical cells; (b) strawberry-like structures; (c) special "sporangia" (SI). 

3. The inflated spherical cells discharge single spores which develop Concho- 
celis colonies. 

4. The interior cells of the strawberry-like structures and unicellular spore-like 
bodies discharged by disintegration of the structure produce germ tubes, and grow 
into Conchocelis colonies. 

5. These Conchocelis colonies produce monosporangia and liberate the mono- 
spores that germinate into normal leafy thalli under short-day conditions. The 
spherical cells and the strawberry-like structures were produced on Conchocelis 
branches at temperatures higher than 18 C. and under long-day conditions (sub- 
dued light). 

6. The spores liberated from sporangia (SI) formed under long-day conditions 

develop into various plantlets: (a) Conchocelis-like plants, globular bodies and 

abnormal blades in long day (14 hours' illumination daily) : (b) plantlets having 

short and root-like filaments, blade-like bodies and massive plantlets in short day 

(9 hours' illumination daily). 

LITERATURE CITED 

DREW, K. D., 1054. Studies in the Bangioidae III. The life-history of Porphyra ninbUicalis 
(L.) Kiitz. var. laciniata (Lightf.) J. Ag. Ann. Bot.. N. S., 18: 183-211. 

KORNMANN, P., 1960. Von Conchocelis zu Porphyra. Hclgolandcr ITiss. Mccrcsuntcrs., 7: 
189-193. 

IWASAKI, H., 1961. The life-cycle of Porphyra tcncra in vitro. Biol. Bull., 121: 173-187. 



<&5 




ENVIRONMENTAL AND TISSUE TEMPERATURES OF SOME 
TROPICAL INTERTIDAL MARINE ANIMALS 

JOHN B. LEWIS 
Bcllairs Research Institute of Me Gill University, St. James, Barbados, W.I . 

Intertidal marine animals may be subjected over short periods of time to a 
broad spectrum of physical conditions. High temperature and desiccation are 
likely to be important factors for tropical intertidal forms. The importance of 
these factors has been noted in determining vertical distribution of gastropods, by 
Broekhuysen (1940). Although they are poikilotherms, it cannot be assumed that 
the body temperatures of intertidal forms are the same as the meteorological values 
of the micro-climate which they inhabit. The effects of insolation and evaporation 
on tissue temperatures of insects and other terrestial arthropods are well known 
(Gunn. 1942; Parry, 1951; Wigglesworth, 1948; Edney, 1951, and many others). 
The body temperatures of intertidal marine animals have, however, received little 
attention, except for a study by Southward (1958) on northern intertidal forms. 
Southward found body temperatures in many cases to be higher than the local 
meteorological values, due to retention of sea water and warming by sunlight. In 
tropical forms one would expect the effect of insolation to be pronounced. The 
preference of certain intertidal animals for shaded areas has been noted by Lewis 
(I960) and Broekhuysen (1940). 

METHODS 

Temperatures were measured with fine wire, copper/constantine. thermocouples 
insulated with lacquer. The e.m.f. was measured against a reference bath of ice 
and water in an insulated Thermos flask at C. The recording instrument was 
a portable potentiometer manufactured by Pye Instruments Ltd., Cambridge, 
England. An accuracy of .25 F. is claimed by the manufacturer. Wet and 
dry bulb air temperatures were measured with a sling psychrometer. 

Three species were selected for study : the barnacle, Tetraclita sqnauiosa, the 
limpet, Fissurella barbadcnsis and the gastropod, Ncrita tesselata. These species 
were abundant at the site chosen, a flat platform of beach rock which is fully ex- 
posed at low tide. Body temperatures were measured by thrusting the thermo- 
couple tip under the operculum of Nerita, through the apical hole of Fissurella and 
between the plates of the carapace of Tetraclita. Temperatures of a black body and 
an inanimate body were also measured. The black body was a piece of soft black 
mastic and the inanimate body, the shell of a Ncrita, stuffed with mastic. The 
deflection of the galvonometer was noted immediately, and each animal was used 
only for a single temperature measurement. 

277 



278 



JOHN B. LEWIS 



RESULTS 

(a) Climate of the intertidal zane 

The observed temperatures of the intertidal zone, recorded in Tables I -VII I, 
clearly indicate that air temperatures measured on the location do not give a com- 
plete assessment of the microclimate. The difference between air temperature 
in a shaded crevice and that of a black body on the rock surface may, for example, 
be as great as 23 F. on a hot sunny day. On a day with no sun the range of 
temperatures is not as great. The difference between air shade temperature and 
black body temperature was only 4 F. on an overcast day. 

Throughout the course of observations the temperatures taken in various places 
measured formed an ascending sequence from wet bulb temperature to black body 
temperature. The lowest temperatures recorded were those of the wet bulb 
thermometer, which was cooled by rapid evaporation. The next lowest was the 
air shade temperature, measured in sheltered crevices close to the water or to a 
wet rock surface. Dry bulb temperatures were slightly higher than air shade 
temperatures but lower than air sun temperatures. The differences between 
these three air temperatures were due to the effects of evaporation and heating 

TABLE I 

Observations of temperatures in the intertidal zone and of tissue temperatures of barnades, 

April 24, 1962, 1200-1300 hours 



Air temp. 
Shade 


Air temp. 
Sun 


Dry 

bulb 


Wet 
bulb 


Sea 

temp. 


Rock 
surf. 


Inanitn. 
body 


Black 
body 


Tissue 

temp. 


82.2 


88.0 


87.2 


73.0 


86.4 


95.1 


101.0 


105.2 


94.8 


85.3 


90.0 


86.5 


73.5 


87.0 


95.5 


98.8 


107.8 


101.0 


84.3 


90.5 


86.3 


72.3 


87.4 


95.3 


98.8 


105.6 


99.6 


86.4 


92.2 


86.0 


72.5 


88.6 


95.1 


99.6 


108.8 


97.4 


85.3 


89.0 


86.0 


72.7 


88.2 


95.2 


99.2 


108.2 


99.2 


84.3 


90.8 


85.0 


73.5 


89.0 


94.3 


102.8 


112.3 


102.4 


Mean 84.6 


90.1 


86.2 


72.9 


87.8 


95.1 


100.0 


107.9 


99.1 



Weather Remarks: No clouds, bright sun, fresh breeze, low tide 1200 hours. 

TABLE II 

Observations of temperatures in the intertidal zone and of tissue temperatures of barnacles, 

May 8, 1962, 1300-1400 hours 



Air temp. 
Shade 


Air temp. 
Sun 


Dry 

bulb 


Wet 
bulb 


Sea 
temp. 


Rock 
surf. 


Inanim. 
body 


Black 
body 


Tissue 
temp. 


87.0 


87.8 


88.0 


76.2 


94.8 


97.4 


105.8 


110.0 


96.8 


85.8 


88.4 


87.5 


76.5 


95.3 


97.2 


101.5 


110.0 


100.8 


89.2 


91.3 


87.5 


76.0 


94.8 


97.2 


102.4 


111.6 


102.0 


86.4 
85.0 


90.8 
91.3 


87.0 
87.0 


76.2 
76.5 


96.1 
95.5 


99.2 
96.4 


103.2 
98.0 


109.6 
106.0 


101.0 

96.8 


Mean 86.7 


89.9 


87.4 


76.3 


95.3 


97.5 


102.2 


109.4 


99.5 



Weather Remarks: 5/10 cloud cover, moderate breeze, low tide 1240 hours. 



TEMPERATURES OF TROPICAL ANIMALS 



270 



TABLE III 

Observations of temperatures, in the intertidal. zone and of tissue temperatures of barnacles, 

July 4, 1962, 1150-1250 hours 



Air temp. 
Shade 


Air temp. 
Sun 


Dry 

bulb 


Wet 
bulb 


Sea 

temp. 


Rock 

surf. 


Inanim. 
body 


Black 
body 


Tissue 
temp. 


79.4 


79.4 


77.3 


77.3 


83.4 


82.0 


82.0 


83.2 


83.8 


79.8 


79.6 


82.6 


80.0 


83.2 


80.5 


82.4 


86.0 


86.4 


81.3 


80.8 


83.3 


81.0 


83.3 


81.8 


82.8 


87.0 


83.8 


82.0 


82.0 


81.2 


78.5 


83.2 


82.0 


84.3 


87.2 


84.0 


81.5 


82.4 


82.0 


78.8 


83.4 


81.5 


84.0 


87.6 


85.0 


81.8 


82.2 


82.0 


79.0 


83.4 


82.0 


85.0 


87.2 


84.3 


Mean 81.2 


81.4 


81.6 


79.1 


83.5 


81.7 


83.7 


86.6 


84.5 



Weather Remarks: 10/10 cloud cover after light rain, very light wind, low tide 1000 hours. 



TABLE IV 

Observations of temperatures in the intertidal zone and of tissue temperatures of gastropods, 

May 21, 1962, 1 100-1200 hours 



Air temp. 
Shade 


Air temp. 
Sun 


Dry 
bulb 


Wet 
bulb 


Sea 

temp. 


Rock 
surf. 


Inanim. 
body 


Black 
body 


Tissue 
temp. 


88.4 


93.6 


86.5 


76.5 


93.0 


95.2 


104.6 


111.2 


99.0 


88.8 


89.6 


86.0 


76.0 


93.0 


95.1 


103.6 


111.4 


96.1 


88.6 


88.4 


89.0 


77.5 


93.0 


99.8 


101.3 


105.4 


92.4 


88.6 


93.0 


88.0 


77.0 


93.0 


99.6 


105.4 


112.0 


95.5 


92.2 


91.0 


90.1 


78.5 


93.0 


99.2 


105.6 


114.0 


95.5 


90.5 


91.0 


88.5 


77.5 


93.0 


104.4 


106.0 


116.8 


94.0 


Mean 89.5 


91.1 


88.0 


77.2 


93.0 


98.9 


104.4 


1 10.1 


95.7 



Weather Remarks: 2/10 cloud cover, fresh breeze, low tide 1102 hours. 



TABLE V 

Observations of temperatures in the intertidal zone and of tissue temperatures of gastropods, 

July 4, 1962, 1150-1250 'hours 



Air temp. 
Shade 


Air temp. 
Sun 


Dry 

bulb 


Wet 

bulb 


Sea 
temp. 


Rock 

surf. 


Inanim. 
body 


Black 
body 


Tissue 
temp. 


79.4 
79.8 

81.3 


79.4 
79.6 
80.8 


77.3 
82.6 
83.3 


77.3 
80.0 } 
81.0 


83.4 
83.2 
83.3 


82.0 
80.5 
81.8 


82.0 
82.4 

82.8 


83.2 
86.0 
87.0 


82.2 
82.2 
82.4 


82.0 


82.0 


81.2 


78.5 


83.2 


82.0 


84.3 


87.2 


83.6 


81.5 


82.4 


82.0 


78.8 


83.4 


81.5 


84.0 


87.6 


83.6 


81.8 


82.2 


82.0 


79.0 


83.4 


82.0 


85.0 


87.2 


82.0 


Mean 81.2 


81.4 


81.6 


79.1 


83.5 


81.7 


83.7 


86.6 


82.6 



Weather Remarks: 10/10 cloud cover after light rain, very light wind, lo\v tide 1000 hours. 



280 



JOHN B. LEWIS 



TABLE VI 

Observations of temperatures in the intertidal zone and of tissue temperatures of gastropods, 

May 8, 1962, 1300-1400 hours 



Air temp. 
Shade 


Air temp. 
Sun 


Dry 

bulb 


Wet 
bulb 


Sea 

temp. 


Rock 
surf. 


Inanim. 
body 


Black 
body 


Tissue 
temp. 


87.0 


87.0 


88.0 


76.2 


94.8 


97.4 


105.8 


110.0 


85.0 


85.8 


88.4 


87.5 


76.5 


95.3 


97.2 


101.5 


110.0 


86.0 


89.2 


91.3 


87.5 


76.0 


94.8 


97.2 


102.4 


111.6 


85.0 


86.4 


90.8 


87.0 


76.2 


96.1 


99.2 


103.2 


109.6 


82.4 


85.0 


91.3 


87.0 


76.5 


95.5 


96.4 


98.0 


106.0 


82.8 


Mean 86.7 


89.9 


87.4 


76.3 


95.3 


97.5 


102.2 


109.4 


84.2 



Weather Remarks: 5/10 cloud cover, moderate breeze, low tide 1240 hours. 

TABLE VII 

Observations of temperatures in the intertidal zone and of tissue temperatures of limpets, 

April 10, 1962, 1300-1400 hours 



Air temp. 
Shade 


Air temp. 
Sun 


Dry 

bulb 


Wet 

bulb 


Sea 
temp. 


Rock 
surf. 


Inanim. 
body 


Black 
body 


Tissue 
temp. 


83.2 


86.0 


82.3 


72.5 


80.0 


85.0 




96.7 


88.8 


81.0. 


84.5 


82.3 


72.5 


80.0 


86.0 




98.8 


86.8 


80.5 


86.8 


82.3 


72.5 


80.0 


85.5 




97.5 


87.2 


Mean 81.6 


85.8 


82.3 


72.5 


80.0 


85.5 




97.6 


87.6 



Weather Remarks: 8/10 cloud cover (measurements made during sunny periods), strong breeze, 
lo\v tide 1346 hours. 

TABLE VIII 

Observations oj temperatures in the intertidal zone and of tissue temperatures of limpets, 

July 4, 1962, 1150-1250 hours 



Air temp. 
Shade 


Air temp. 
Sun 


Drv 
bulb 


Wet 
bulb 


Sea 
temp. 


Rock 
surf. 


Inanim. 
body 


Black 
body 


Tissue 
temp. 


79.4 


79.4 


77.3 


77.3 


83.4 


82.0 


82.0 


83.2 


85.5 


79.8 


79.6 


82.6 


80.0 


83.2 


80.5 


82.4 


86.0 


85.5 


81.3 


80.8 


83.3 


81.0 


83.3 


81.8 


82.8 


87.0 


86.0 


82.0 


82.0 


81.2 


78.5 


83.2 


82.0 


84.3 


87.2 


85.8 


81.5 


82.4 


82.0 


78.8 


83.4 


81.5 


84.0 


87.6 


86.4 


81.8 


82.2 


82.0 


79.0 


83.4 


82.0 


85.0 


87.2 


86.0 


Mean 81.2 


81.4 


81.6 


79.1 


83.5 


81.7 


83.7 


86.6 


85.9 



Weather Remarks: 10/10 cloud cover after light rain, very light wind, low tide 1000 hours. 



by the sun. Air in shaded crevices received no sunlight and was cooled by 
evaporation troni the water surface. Dry bulb temperatures were taken in the 
shade, while the air sun temperatures, measured with thermocouples, were not 
subject to the above cooling influences. 



TEMPERATURES OF TROPICAL ANIMALS 281 

Sea temperatures varied considerably and were observed to be above or below 
air sun temperature, depending upon the height of the tide. This was due to the 
fact that measurements were made in tide-pools which, at very low tides, were 
heated up by the sun and at higher water levels received periodic flushing from 
the sea. The temperatures of the rock surface were always several degrees above 
air temperatures, but varied with the degree of wetness. 

The two extraneous elements introduced, the inanimate body and the black 
body, both produced temperatures higher than those measured elsewhere. The 
black body absorbed the maximum amount of radiation and had invariably the 
highest observed temperature. The observed temperatures of the inanimate body 
were lower than the black body but consistently higher than air temperatures. The 
source of heating above air temperature was presumably mainly the incoming 
radiation. 

(b) Tissue temperatures 

The results of observations of temperatures relating to the barnacles are shown 
in Tables I, II and III. Under conditions of full sunlight and a fresh breeze the 
body temperatures of the barnacles rose 9 F. above that of the air sun temperature. 
Body temperatures were also 4 F. higher than rock surface temperatures, but 
below those of the inanimate body and black body of 0.9 and 8.8 F., respectively. 

On a partly cloudy day with moderate breeze, tissue temperatures of barnacles 
were above air sun temperatures by 9.4 and above rock surface temperatures by 
2 F. Tissue temperatures were 2.7 lower than inanimate body and 9.9 F. 
lower than the temperatures of the black body. 

Table III contains the observations on a completely overcast day. Tempera- 
tures were recorded shortly after a light rain. On this day tissue temperatures 
\vere only 3.1 above air sun temperatures and 2.8 F. above the rock surface. 
The temperatures of the inanimate body were 0.8 lower than tissue temperatures. 
Black body temperatures were only 2.1 F. higher than tissue temperatures. 

On all three days body temperatures of barnacles were above air temperature. 
The differences were greatest on sunny days and considerably less on an overcast 
day. It is apparent, then, that the animals absorbed heat from incoming radiation. 
Body temperatures on sunny days rose nearly as high as the temperature of an 
inanimate body of the same size and color but not as high as a black body which 
absorbs the maximum amount of radiation. 

The results of observations relating to the gastropod are shown in Tables IV, V 
and VI. On a day of full sunlight with a fresh breeze, tissue temperatures \vere 
4.6 above the air sun temperature but below those of the rock surface, inanimate 
body and black body by 3.2. 8.7 and 14.4 F., respectively. 

On a partly cloudy day, body temperatures of the gastropods in shaded crevices 
were below shade air temperatures by 2.5 and sun air temperatures by 5.7 F. 
The body temperatures were far below that of an inanimate body by 18 F. 

On a cloudy overcast day the difference between tissue temperature and en- 
vironmental temperature was much less marked. Air sun temperature was 0.6 
lower than tissue temperature, and inanimate body and black bodv temperatures 
were only 1.1 and 4 F. higher than tissue temperatures. 

Thus, on the two sunny days the tissue temperatures of Xcritu were very close 



JOHN B. LEWIS 

to or below those of the ambient air temperatures. On neither day did they rise 
as high as the rock surface or inanimate body temperatures. While those specimens 
in sheltered crevices absorbed little of the incoming radiation, it is apparent that 
those exposed to the sun do not heat up in spite of radiation absorption. 

The results of observations relating to limpets are shown in Tables VII and 
VIII. On a partly cloudy day, temperatures were recorded during sunny periods. 
Tissue temperatures were 1.8 above air sun temperatures and 2.1 F. above rock 
surface temperatures. Black body temperatures were 10 above tissue tem- 
peratures. 

On an overcast day tissue temperatures were above air sun temperatures, rock 
surface and inanimate body temperatures by 4.5, 4.2 and 2.2 F., respectively, 
but below that of the black body by 0.7 F. 

On both days, tissue temperatures were above air temperatures and rock sur- 
face temperatures, but below those of the black body. It is apparent that this 
species is absorbing incoming radiation, but on a sunny day the body heat rises 
only slightly above air temperature. 

DISCUSSION 

It is evident from the above results that tissue temperatures of animals living in 
the intertidal zone are not the same as the ambient air temperatures during ex- 
posure. Southward (1958) has attributed the rise in body temperature above sea 
and air temperatures of intertidal animals at Plymouth to warming by sunlight. 
Clearly this effect will be much more marked in the tropical forms. 

Our results indicate, however, that intertidal animals do not absorb radiation as 
do inanimate bodies or black bodies. Thus, on a sunny day the tissue tempera- 
tures of barnacles are below black body temperatures by 8.8, the gastropod Nerita 
by 14.1 and limpets by 10 F. Similarly, they were all below the temperature of 
an inanimate body. 

It is thus apparent that there is some factor acting to reduce the heating effect 
of the sun. The most obvious cooling mechanism is that of evaporation. The 
effects of evaporation from the cuticle of insects have already been noted. In 
some insects evaporation may cool the body temperature several degrees C. below 
the air temperature (Gunn. 1942). It thus seems likely that a similar mechanism 
may be operative on intertidal animals. 

In barnacles the plates of the carapace are tightly closed during exposure. 
There thus seems little opportunity for evaporation to occur in this species after 
the initial drying of the external shell. In Nerita, however, the foot attached to 
the substrate remains moist during exposure and the aperture of the shell is never 
pressed tightly downwards. When a specimen is lifted off the rock, the operculum 
closes and a drop of water remains in the aperture. This drop feels distinctly cool 
when placed on the skin. The temperature of the drop measured by a thermo- 
couple did in fact show it to be identical with body temperature of the specimen. 
In limpets the shell is not firmly pressed against the substrate, so that evaporation 
can take place around the mantle. Furthermore, in Fissitrclla there is an apical 
hole through which evaporation can also take place. 

In barnacles there is apparently the least opportunity for evaporation to take 
place, and observed tissue temperatures of this species were closer to those of the 



TEMPERATURES OF TROPICAL ANIMALS 

TABLE IX 

Water loss in Nerita during exposure on a cloudy day 



283 



Spec. no. 


Wt. before 
exposure 


Wt. after 1-hr, 
exposure 


Wt. after 2-hr, 
exposure 


Wt. after 4-hr, 
exposure 


Wt . after 6-hr. 
exposure 


\V;tter loss 
f, lir. 


1. 


24.056 


24.051 


24.046 


24.035 


24.026 


0.030 


2. 


23.947 


23.938 


23.930 


23.915 


23.906 


0.041 


3. 


25.713 


25.709 


25.703 


25.692 


25.683 


0.030 


4. 


21.568 


21.561 


21.557 


21.551 


21.545 


0.023 


5. 


23.774 


23.763 


closed 








6. 


22.429 


22.416 




22.398 


22.391 


0.038 


7. 


21.507 


21.504 


21.501 


21.493 


closed 





inanimate body and black body, than the tissue temperatures of the limpets and 
gastropods. 

Low body temperatures were most striking in the gastropod Nerita. In this 
species the tissue temperatures on a sunny day were lower in relation to environ- 
mental temperatures than in the other two species. If evaporative cooling is more 
effective in this species, then there will be a considerable loss of water during ex- 
posure, and hence a loss in body weight. Gowanloch (1926) has demonstrated 
heavy water loss of northern gastropods during exposure. 

The losses in weight of animals exposed in full sunlight and on a cloudy day 
are shown in Tables IX and X. Each animal was placed on a glass dish with the 
operculum open and the foot attached to the glass. The glass with animal was 
weighed to the nearest 0.001 gram before exposure, and again at hourly intervals. 
Only those animals which kept their opercula open during the exposure period 
were considered. 

The results of Tables IX and X indicate loss of water both in sunlight and 
on a cloudy day. There is a marked difference in the rate of evaporation, how- 
ever. On a cloudy day there was a loss of less than 0.05 gm. of water in six hours. 
On a sunny day the loss in only two hours was between 0.05 and 0.1 gm. It would 
appear, then, that this species has an effective control over its body temperature 
by losing water during the exposure period in the intertidal zone through 
evaporation. 



TABLE X 
Water loss in Nerita during exposure on a sunny day 



Spec. no. 


Wt. before 
exposure 


Wt. after 1-hr, 
exposure 


Wt. after 2-hr, 
exposure 


Water l<>~. 
1 lir. 


Water 1<>". 
2 hr. 


1. 


29.161 


29.101 


closed 


0.050 




2. 


22.916 


22.882 


22.863 


0.034 


0.053 


3. 


24.167 


24.120 


24.100 


0.047 


0.067 


4. 


23.607 


23.540 


closed 


0.067 




5. 


21.554 


21.498 


21.471 


0.056 


0.083 


6. 


20.727 


20.703 


closed 


0.024 





284 JOHN B. LEWIS 

The difference among the three species studied, in resisting increase in body 
temperatures during exposure, suggests a relation to the vertical distribution of the 
three forms. At the site chosen, Fissure I la and Tetraclita live at approximately 
the mid-tide level, while Nerita, which has the greatest control over its body tem- 
perature, is found at high tide (Lewis, 1960). Thus, the ability of intertidal 
forms to regulate their body temperatures may, in the tropics, have a bearing on 
intertidal zonation. 

SUMMARY 

1. Body temperatures of three common intertidal animals were measured with 
fine wire thermocouples. 

2. Body temperatures were observed to be considerably above ambient air 
temperatures on a hot sunny day. 

3. The difference between body temperatures and ambient air temperatures 
was less marked on a cloudy day. 

4. Evaporative cooling was apparently the mechanism for lowering body tem- 
peratures. This process was most effective in the gastropod, Nerita. 

5. A relationship between ability to regulate body temperature by evaporation 
and intertidal zonation is suggested. 

LITERATURE CITED 

BROEKHUYSEX, C. J., 1940. A preliminary investigation of the importance of desiccation, tem- 
perature and salinity as factors controlling the vertical distribution of certain marine 
gastropods in False Bay, South Africa. Trans. Roy. Soc. S. Africa, 28: 255-292. 

EDNEY, E. B., 1951. The evaporation of water from woodlice and the millipede Gloincris. 
J.E.i-p. Bio!., 28: 91-115. 

GOWANLOCH, J. N., 1926. Contributions to the study of marine gastropods. II. The intertidal 
life of Bnccitunn undatnin, a study in non-adaption. Contr. Canad. Biol., N.S., 3: 
169-177. 

GUNN, D. L., 1942. Body temperatures in poikilothermal animals. Biol. Rcrs., 17: 293-314. 

LEWIS, J. B., 1960. The fauna of rocky shores of Barbados, West Indies. Canad. J. Zool, 
38: 391-435. 

PARRY, D. A., 1951. Factors determining the temperatures of terrestial arthropods in sunlight. 
J.Exp.Biol.,28: 445-462. 

SOUTHWARD, A. J., 1958. Note on temperature tolerances of some intertidal animals in relation 
to environmental temperatures and geographical distribution. /. Mar. Biol. Assoc., 
37 : 49-66. 

WIGGLESWORTH, V. B., 1948. The insect cuticle. Bio}. Revs., 23: 408-451. 



ACID AND ALKALINE PHOSPHATASE CHANGES ASSOCIATED 
WITH FEEDING. STARVATION AND REGENERATION 

IN PLANARIANS 1 

PAUL J. OSBORNE AND A. T. MILLER, JR. 

Department of Biology, Lynelihitr;/ Collet/e. Lynehhitri/. I'irt/iuia, and I >ef>urfinent of I'livsioloi/y. 

University of North Carolina, Chapel Hill, A'. C. 

The acid phosphatase activity in planarian gastrodermal cells has been reported 
to increase almost immediately following the ingestion of food (Rosenbaum and 
Rolon, 1960; Jennings, 1962b). Acid phosphatase has also been implicated in the 
early stages of digestion in the food vacuoles of paramecia (Miiller and Toro, 1962) 
and of amebae (Miiller, Toth and Toro, 1962). These results suggest an analogy 
between the food vacuoles of lower animals and the lysosomes ( de Duve, 1961) of 
higher animals. Lysosomal enzymes are thought to be involved not only in the 
digestion of exogenous materials but also in the autolytic degradation of tissues 
in such processes as developmental involution and metamorphosis. It seems 
likely, therefore, that acid hydrolases may play a similar role in the gradual dis- 
appearance of digestive and reproductive organs during prolonged starvation in 
planarians, as well as in the provision of raw materials for the early stages of 
regeneration following transection, before the ingestion of exogenous food stuffs 
can be resumed. 

Less is known about the significance of alkaline phosphatase than of acid 
phosphatase, although its frequent localization in absorptive epithelia in higher 
forms is suggestive of a role in the phosphorylation of certain compounds prior 
to their transport across membranes. High levels of alkaline phosphatase activity 
have been observed in several organs of planarians, including protonephridia 
(Danielli and Pantin, 1950), resting neoblasts (Pedersen, 1959) and nervous and 
muscular tissues (Gazso, Torok and Rappay, 1961). Jennings (T962b) has re- 
cently reported the appearance of alkaline phosphatase in planarian gastrodermal 
cells several days following food ingestion, and he suggests that it may be con- 
cerned with the release of energy needed for .secretion of the various digestive 
enzymes and for the absorption of the products of digestion from the food vacuoles. 

The object of the present work was to extend the above-mentioned observations 
to later stages of digestion as they are succeeded by starvation effects, and to the 
tissue reconstruction involved in regeneration following transection. 

MATERIALS AND METHODS 

All observations were made on specimens of Duc/esia tii/rhia obtained from the 
Carolina Biological Supply Company. Krozen sectinis were used instead of paraf- 

1 Supported by grants from the National Institutes of Health (C-3996 and A-1699) and the 
National Science Foundation (NSF-GB-223 and NSF-GA-282). 

285 



286 




PAUL J. OSBORNE AND A. T. MILLER, JR. 

"j> IO, H * ;'? 




la 



/y. v)* 

> &>'/ '2/V 












3 




. 

^ 









-," 
:*? 



*. I , . - . .TV ? JT * ^-^1 ' **, * % - - if ' 



, x^,^-.^ 



FIGURE la. Acid phosphatase activity in gastrodermal cells 24 hours after ingestion of 
cooked egg yolk. Arrow points to a gastrodermal cell containing two food vacuoles. 900 X. 

FIGURE Ib. Target form of gastrodermal cell, with acid phosphatase activity in cell mem- 
brane and in central core, two days after ingestion of cooked egg yolk. 1350 X. 

FIGURE 2. Eight days after ingestion of cooked egg yolk. The gastrodermal cells (solid 
arrows) are free of acid phosphatase activity and the gut region is being invaded by neoblasts 
(dotted arrows ) with high acid phosphatase activity. 675 X. 

FIGURE 3. Thirty days after ingestion of cooked egg yolk. The intestinal epithelium is 
densely infiltrated by neoblasts with moderately high acid phosphatase activity. 300 X. 

FIGURE 4. Alkaline phosphatase activity in food vacuoles 24 hours after ingestion of cooked 
egg yolk. 300 X. 



PHOSPHATASES IX PLAXA K I ANS 287 

fill-embedded sections, despite the hetter cvtolngical preservation <>!" the latter, 
because of the desire to preserve en/ynie activity. This is especially important in 
the case of acid phosphatase, which undergoes considerable inactivation even wlien 
low-melting-point paraffin is used. 

After a ten-day period of starvation (to eliminate the enzyme changes associ- 
ated with previous feeding) the worms were allowed to feed on cooked egg yolk 
until satiated, and were then removed to fresh medium containing no food. At 
varying intervals after feeding, worms were slowly chilled to 4 C. and then fixed 
for 12-24 hours at 4 C. in W r / f neutral formalin containing \ c / ( calcium chloride. 
The fixed specimens were rinsed for 1-2 hours in distilled water at 4 C., then 
blotted and embedded in W% gelatin, in the following manner. Melted gelatin 
was layered in the bottom of a small beaker and chilled until firm. Specimens 
were placed on top of the solidified gelatin, covered with a layer of melted gelatin 
and the beaker placed in a freezer until the gelatin had solidified. The gelatin 
was then cut into blocks, each containing a single worm ; the blocks were mounted 
on cryostat object holders and frozen with dry ice or Freon. Sections were cut 
at 8 microns in a Pearse cryostat, mounted on chilled slides and air-dried. Acid 
and alkaline phosphatase activities were visualized bv the methods of Gomori 
(1952). 

For the studies on phosphatase changes during regeneration, planarians were 
starved for 10 days, the pharynx was carefully removed with dissecting needles, and 
the worm was cut transversely into equal halves which were allowed to regenerate 
in food-free medium. Each half regenerated its missing structures, including a 
pharynx, and the results to be described were identical in organisms regenerating 
from anterior and posterior halves. 

RESULTS 

Twenty-four hours after the ingestion of cooked egg yolk there was a moder- 
ately high level of acid phosphatase activity in the gastrodermal cells (Fig. la). 
This confirms the reports of Rosenbaum and Rolon (1960) and of Jennings 
(1962b). The enzyme reaction product at this time was in the form of small 
granules which gradually increased in size and numbers until they filled the 
entire cell by the second day after ingestion of food. At this time some of the 
gastrodermal cells had a "target" appearance, with intense acid phosphatase ac- 
tivity in the cell membrane, separated by a clear zone from the central core of 
enzyme activity (Fig. Ib). 

Acid phosphatase activity in the gastrodermal cells diminished rapidly after the 
second day following food ingestion ; by the fourth day many of the cells had no 
demonstrable activity, though minimal activity persisted in a few cells up to seven 
days. Coincident with the decline in acid phosphatase activity in the gastrodermal 
cells there was a striking increase in the activity of this enzyme in the neoblasts. 
By the eighth day there was a definite invasion of the gut region by acid phospha- 
tase-rich neoblasts (Fig. 2). Thirty days after the ingestion of food the walls of 

FIGURE 5. Alkaline phosphatase activity in pharynx, nerve fibers and glands after 60 days' 
starvation. 115 X. 

FIGURE 6. Regenerating planarian 6 days after transaction (longitudinal section). The 
regenerating pharynx (a) is densely infiltrated with neoblasts having acid phosphatase activity, 
as are the original gut (b) and the regenerating gut (c). 75 X. 



288 



PAUL J. OSBORNE AND A. T. MILLER, JR. 




4k * 
JH* 






I 



u 

8 



a 








a 



10 




U 




12 



FIGURE 7. Higher magnification of cells of regenerating pharynx in Figure 6, showing 
neoblasts with acid phosphatase activity. 675 X. 

FIGURE 8. Phase contrast photograph of neoblasts in "inactive" region ; edge of zone of 
regenerating gut at top of field. 675 X. 

FIGURE 9. Light microscope photograph of same field as Figure 8. Note the absence of 
acid phosphatase activity in all the neoblasts except those near the "active" region, a, zone of 
regenerating gut : b, inactive zone. 675 X. 

FIGURE 10. Gradient of acid phosphatase activity in neoblasts migrating (from right to 
left) to region of regenerating gut. a and b as in Figure 9. 675 X. 



PHOSPHATASES IX PLANARIANS 289 

the- gut diverticnli were densely infiltrated by ncoblasts in which the level <>f acid 
phosphatase activity was beginning to decline ( Fig. .}). Coincident with the 
changes in the gut region, high levels of acid phosphatase activity were observed in 
the mucous glands and ventrolateral slime tracts and in the neoblasts in the 
pharynx. 

The ingestion of cooked food was followed by the appearance of alkaline phos- 
phatase activity in the forming food vacuoles (Fig. 4). After reaching a peak one 
to two days after food ingestion, the enzyme activity diminished somewhat more 
slowly than did that of acid phosphatase and was absent by the eighth day. Al- 
kaline phosphatase activity in nerve fibers, mucous glands and pharynx was not 
affected by feeding and was undiminished after periods of starvation as long as 60 
days (Fig. 5). 

Preliminary studies have been made on phosphatase changes associated with 
regeneration following transection. On the sixth day after transection the re- 
generating pharynx was heavily infiltrated with neoblasts which had a moderately 
high level of acid phosphatase activity (Figs. 6 and 7). Figure 6 also shows an 
infiltration of both the original and the regenerating gut with neoblasts rich in acid 
phosphatase, suggesting that these cells were participating both in the degradation 
of the original gut (starvation effect) and in the reconstitution of the gut in the 
regenerating portion of the worm. The neoblasts at some distance from the sites 
of organ destruction or reconstitution ("resting" neoblasts) were lacking in acid 
phosphatase activity. As they migrated toward these sites they appeared to acquire 
progressively increasing amounts of acid phosphatase (Figs. 8-10). 

The regenerating pharynx was also rich in alkaline phosphatase, predominantly 
localized in the neoblasts (Figs. 11 and 12). 

DISCUSSION 

The changes in acid and alkaline phosphatase activities associated with feeding, 
starvation and regeneration in planarians suggest that these enzymes are intimately 
involved in the processes of nutrition, growth and repair. Little is known, how- 
ever, about the stimuli to enzyme induction and the exact chemical reactions which 
are catalyzed by these enzymes in vivo. 

Rosenbaum and Ro 1 on ( 1960) reported the appearance of acid phosphatase 
activity in the gastrodermal cells of Dngesia dorotocephala and Dugesia tif/rina 
within five minutes after the feeding of cooked liver, and they identified this with 
the formation of food vacuoles. Maximum acid phosphatase activity was present 
in all the food vacuoles 24 to 48 hours after feeding ; the level of enzyme activity 
began to decline three days after feeding and was absent one week after feeding. 
Similar changes were observed in aminopeptidase and /2-glucuronidase activities, 
although intracellular localization of the latter enzyme was not achieved. Miiller, 
Toth and T<">r<"> ( 1962 ) described a similar relation of acid phosphatase and non- 
specific esterase activity to the food vacuole cycle in Amoeba protcus. In unfed 
amebae, fine granules giving an intense acid phosphatase reaction (lysosomes?) 

FIGURE 11. Regenerating pharynx 6 days after transection. The infiltrating neoblasts are 
rich in alkaline phosphatase. 75 X. 

FIGURE 12. Higher magnification of regenerating pharynx in Figure 11, showing alkaline 
phosphatase in neoblasts. 675 X. 



290 PAUL J. OSBORNE AND A. T. MILLER, JR. 

were distributed at random. Recently-formed food vacuoles showed only moderate 
en/ymc activity but strongly staining grannies were observed around these early 
vacuoles, often forming large aggregates. The authors suggest that these granules 
may be "enzyme carriers." Vacuoles containing still living or dying prey (Tetra- 
hyuiena pyriformis) showed no demonstrable increase in acid phosphatase activity 
("ingestion vacuoles"), but a sharp rise in enzyme activity occurred when the food 
organisms had been killed and most of the water had disappeared from the vacuoles 
("digestion vacuoles"). Parallel changes occurred in non-specific esterase activity. 
The authors relate the increased activities of the two enzymes to the processes of 
intracellular digestion of food and suggest that ameba digestion vacuoles can be 
considered as lysosomes. Presumably they would be representatives of the type 
called "lysophagosomes" by de Duve (1961). 

Our results confirm the recent report of Jennings (1962b) of a sequential 
appearance of acid and alkaline phosphatase activities in the gastrodermal cells of 
planarians following food ingestion. The apparently greater intensity and longer 
persistence of acid phosphatase activity in our preparations is probably due to 
better enzyme preservation in frozen sections than in paraffin-embedded material. 
There is general agreement that acid phosphatase and other acid hydrolases are 
involved in the initial stages of digestion in food vacuoles (Rosenbaum and Rolon, 
1960; Miiller et al., 1962; Jennings, 1962b) but the function of the alkaline phos- 
phatase which appears some hours after the beginning of digestion is unknown. 
Jennings (1962b) suggests that it "may be concerned in the release of energy 
needed for secretion of the various enzymes and the absorption of the products of 
digestion from the vacuoles." In studies on the rhynchocoelan, Linens ntbcr, 
Jennings (1962a) reported the appearance of intense alkaline phosphatase activity 
in the luminal margins of the gut cells immediately after feeding, and he postulated 
a role of the enzyme in the process of phagocytosis of food. No such immediate 
response of alkaline phosphatase to food ingestion has been observed in planarians. 

Prolonged starvation in planarians is characterized by the resorption of certain 
structures, notably the digestive and reproductive systems, and by the persistence 
of other more immediately essential structures, such as the pharynx, nervous sys- 
tem, protonephridia and lateral mucous glands. Survival for periods in excess 
of 60 days without food has been observed, and interesting changes in phosphatase 
activity occur during starvation. The earliest of these changes is a striking in- 
crease in acid phosphatase activity in the neoblasts, especially those surrounding 
the gut region, which is clearly in evidence on the fourth day following feeding. 
In the succeeding days the neoblasts infiltrate the gut where their hydrolytic 
enzymes presumably participate in the autolytic degradation of the gastrodermal 
cells in the later stages of starvation. 

The sequence of changes in alkaline phosphatase activity during starvation is 
quite different. Following the decline in activity in the gastrodermal cells, leading 
to disappearance of the enzyme about eight days following feeding, there is no 
recurrence as in the case of acid phosphatase. However, the alkaline phosphatase 
activity of those structures which do not undergo resorption during starvation 
seems to be unaffected by feeding and starvation. Thus, even after 60 days starva- 
tion, there is intense alkaline phosphatase activity in the pharynx, the nerve fibers 
and the lateral mucous glands, i.e., those structures which by preserving the 



PHOSPHATASES IX PLANARIANS 291 

capacity for locomotion and lor ingestion <>| food, lavor the survival nl the 
organism. 

The preliminary observations on regenerating planarians suggest that the 
neoblasts play a role lioth in the (probably) degenerative changes in the gut of the 
original segment and in the reconstitution of organs in the regenerating segment. 
In both cases the neoblasts in these sites have a high level of acid phosphatase ac- 
tivity. Since the "resting" neoblasts show no acid phosphatase activity (Figs. 
810), this must represent enzyme induction and a participation of acid phosphatase 
in both degradative and regenerative processes. The neoblasts in the sites of organ 
regeneration have intense alkaline phosphatase activity also, but the functional 
significance of this observation is uncertain, since resting neoblasts also show 
alkaline phosphatase activity (Pedersen, 1959). 

The results of the present study implicate acid phosphatase in a variety of deg- 
radative processes, presumably autolytic in nature, all of which results in the con- 
version of endogenous or exogenous compounds into building materials for main- 
tenance and repair. An apparent discrepancy between our results and those of 
earlier studies on the histological changes during starvation in planarians requires 
some comment. Willier ct al. (1925) reported that the cells of the intestinal 
epithelium show practically no change until after six weeks of starvation, when 
they begin to undergo degeneration with reduction in size. Our observation of 
intense acid phosphatase activity in the region of the intestinal epithelial cells after 
much shorter periods of starvation suggests that the enzymatic phase of autolytic 
degradation may begin much earlier, but that this does not lead immediately to 
histologically demonstrable degeneration of the cells. The participation of acid 
phosphatase in the reconstitution of organs in regenerating planarians suggests the 
simultaneous occurrence of degradative and synthetic reactions, the one perhaps 
providing the raw materials for the other. The functional significance of the 
alkaline phosphatase changes is unknown. Its presence in the neoblasts in regions 
of regeneration may reflect a participation in organogenesis (Junqueira, 1950; 
Vorbrodt, 1958), but this is conjectural in view of the occurrence of alkaline 
phosphatase in resting neoblasts. 



The senior author is indebted to Dr. E. Ruffin Jones, Jr., Biology Depart- 
ment, University of Florida, for aid in earlier studies (unpublished) on the hy- 
drolytic enzymes of planaria. 

SUMMARY 

1. The ingestion of cooked food by starved planarians is followed by the ap- 
pearance of high levels of acid, and later of alkaline phosphatase activities in 
the gastrodermal cells. The enzyme activities decline gradually, to disappear after 
7-10 days. Beginning about 4 days after feeding, the gut region is progressively 
invaded by neoblasts with high acid phosphatase activity. Intense alkaline phos- 
phatase activity persists in certain "essential" structures (nerve fibers, gland cells, 
protonephridia) even after periods of starvation as long as 60 days. 

2. During regeneration following transection. neoblasts rich in acid phosphatase 
invade both the regenerating organs and the degenerating structures of the original 



292 PAUL J. OSBORNE AXU A. T. MILLER, JR. 

segment. This represents en/vine induction, since resting neohlasts show no acid 
phosphatase activity. Xeohlasts witli alkaline phosphatase activity are abundant 
in the regions of regeneration, Init the significance of this observation is uncertain 
since alkaline phosphatase activity also characterizes resting neoblasts. 

3. It is suggested that the lysosomal acid hydrolases (typified by acid phospha- 
tase) are involved not only in the early stages of digestion in the food vacuoles. but 
also in the autolysis of dispensable organs during starvation and in the tissue 
breakdown which precedes regeneration in transected planarians. 

LITERATURE CITED 

DANIELLI, J. F., AND C. F. A. PANTIN, 1950. Alkaline phosphatase in protonephridia of ter- 
restial nemertines and planarians. Quart. J. Micr. Sci., 91 : 209-214. 

DE DUVE, C., 1961. In: Biological Approaches to Cancer Chemotherapy, New York, Academic 
Press, Inc., pp. 101-112. 

GAZSO, L. R., L. J. TOROK AND GY. RAPPAY, 1961. Contributions to the histochemistry of the 
nervous system of planarians. Acta Biol. Acad. Sci. Hunt/., 11: 411-428. 

GOMORI, G., 1952. Microscopic Histochemistry. University of Chicago Press, Chicago, Illinois. 

JENNINGS, J. B., 1962a. A histochemical study of digestion and digestive enzymes in the 
rhynchocoelan Linens rubcr ( O. F. Muller). Biol. Bull., 122: 63-72. 

JENNINGS, J. B., 1962b. Further studies on feeding and digestion in triclad Turbellaria. Biol. 
Bull., 123: 571-581. 

JUNQUEIRA, L. C. U., 1950. Alkaline and acid phosphatase distribution in normal and regenerat- 
ing tadpole tails. /. Anat., 84: 369-373. 

MULLER, M., AND I. TOKO, 1962. Studies on feeding and digestion in protozoa. III. Acid 
phosphatase activity in food vacuoles of Paniincciiint multimicronucleatum. J. Proto- 
col., 9: 98-102. 

MULLER, M., J. TOTH AND I. TORO, 1962. Studies on feeding and digestion in protozoa. IV. 
Acid phosphatase and nonspecific esterase activity of food vacuoles in Amoeba protcus. 
Acta Biohgica Acad. Sci. Hung., 13: 105-116. 

PEDERSEN, K. J., 1959. Cytological studies on the planarian neoblast. Zcitschr. f. Zellforsch., 
50: 799-817. 

ROSENBAUM, R. M., AND C. I. RoLON, 1960. Intracellular digestion in the phagocytic cells of 
planaria. Biol. Bull., 118: 315-323. 

VORBRODT, A., 1958. Histochemically demonstrable phosphatase and protein synthesis. .r/>. 
Cell Res., 15: 1-20. 

WILLIER, B. H., L. H. HYMAN AND S. A. RIFENBURGH, 1925. A histochemical study of intra- 
cellular digestion in triclad flatworms. /. Morphol. and PhysioL, 40: 299-340. 



STUDIES ON THE HEXAPOD NERVOUS SYSTEM. VI. VENTRAL 

\ERVE CORD SHORTENING; A METAMORPHIC PROCESS 

IX GALLERIA MELLONELLA (L.) (LEPIDOPTERA. 

PYRALLIDAE) 1 

RUDOLPH L. PIPA 

Department of Entomology and Parasitology, University of California, Berkeley 7, California 

One of many spectacular occurrences during the metamorphosis of holometabo- 
lous insects is the transformation of the larval central nervous system into that of 
the adult. An account of the gross features of the change as it occurs in certain 
Lepidoptera was given in the last century (Newport, 1832, 1834). The salient 
characteristics noted include : a concentration of the mesothoracic, metathoracic, and 
first two abdominal ganglia ; a shortening of connectives between the brain and 
subesophageal ganglion, and between the prothoracic and mesothoracic ganglia. In 
Papilio nrticae L., under the ambient temperatures employed, these changes were 
realized during the first 58 hours after pupation. This contrasted with observations 
on Sphinx lit/itstri L., where nearly six months of developmental arrest (diapause) 
intervene between the most active phases of the phenomenon. 

Brandt's (1879) extensive comparative study partly confirmed Newport's find- 
ings, but called attention to departures from the above pattern of reorganization in 
different lepidopteran species. He also depicted a fusion of the last three abdomi- 
nal ganglia, an event undescribed by Newport. 

The prospect of using the greater wax moth, Galleria incllonclla (L.), in an 
extended analysis of neurometamorphosis has prompted the following investiga- 
tion of the shortening process. Concomitant cytological features will be reported 
in a future communication. 

METHODS 

Stock cultures of Galleria were maintained by placing several adult males and 
females in a screen-topped gallon jar containing a larval diet mixture of 1200 ml. 
Gerber Mixed Cereal, 100 ml. honey. 100 ml. glycerin, and 50 ml. water. These 
were kept in darkness in a constant temperature cabinet (32-35 C.). Each cul- 
ture was divided at least once and replenished with fresh diet to circumvent un- 
desired effects of crowding. Last-instar larvae were gathered from cocoons spun 
along the sides of the containers. 

To ascertain deviations in nerve cord length due to variations in body size, last- 
instar larvae were separated into two groups. Individuals of one group weighed 
170-190 mg., those of the other, 200-250 mg. Because values obtained from the 
two groups were in reasonable accord, they have been combined for statistical 
treatment. 

Ten to 15 larvae were routinely placed in a 105 X 20 mm. plastic Petri-type 
culture dish, provided with cardboard tents, and maintained at a uniform tempera- 

1 This study was supported by National Institutes of Health Grant No. B-3845. 

293 



294 RUDOLPH L. PIPA 

lure (30-33 C.). I>y the: end of a day ihe majority had enclosed themselves 
within cocoons beneath the tents. 'I he ends of each cocoon were opened at 8 12- 
hour intervals to permit inspection for signs of impending pupation, which are 
descrihed below. Larvae demonstrating such signs were removed from their 
cocoons, and placed in separate dishes also kept at 30-33 C. The time during 
which they cast their exuviae, thus revealing "white" (untanned) pupae, was 
known precisely, or else approximated to within three hours. 

Ventral nerve cords in desired stages of development were exposed by dorsal 
dissections of living insects pinned to the bottoms of Syracuse watch glasses partly 
filled with paraffin. The animals were immersed in Yeager's (1939) insect mag- 
nesium saline during the operation. Interganglionic connectives were measured 
to the nearest 0.1 mm. at a magnification of 25 diameters by means of an ocular 
micrometer inserted in a dissecting microscope (Wild-Greenough). Particular 
care was taken to minimize stretching the cords beyond their rest lengths. Meas- 
urements were routinely completely 40-50 minutes after dissection was initiated. 
Comparative measurements made on the same connectives at the beginning and 
end of this period usually differed by less than 10%. 

Measurements of nerve cords from pupae and adults reared from last-instar 
larvae kept at 30-33 C. and 70% relative humidity (R. H. ) over KOH (Buxton, 
1931) did not differ significantly from those made on individuals of comparable 
developmental stage maintained at the same temperature, but at ambient R. H. 
Consequently, no attempt was made to control this factor. 

DEVELOPMENTAL STAGES 

All systems of a multicellular organism do not complete differentiation simul- 
taneously. This, and the fact that developmental transitions are gradual, rather 
than saltatory, necessitates that selection and definition of stages be arbitrary. 
One criterion which might be used is the ability of the larva to spin a cocoon. 
As indicated by Piepho (1950) and Wiedbrauck (1955) this appears to be under 
hormonal control, a decrease in juvenile hormone concentration, such as occurs 
at the outset of pupation (Gilbert and Schneiderman, 1961; Williams, 1961), 
resulting in cessation of spinning. 

During the present investigation an event which precedes total loss of spinning 
activity was detected. This involves the ability of last-instar larvae to protract 
and retract their anal legs, or postpedes, when these are touched. Because the 
reflex is lost 19-27 hours prior to ecdysis, it has served as a useful external sign 
of incipient pupation. The change does not seem to interfere with locomotion, nor 
with the larva's ability to right itself when placed in a supine position. 

Last-instar larvae which have constructed cocoons, which can move their 
postpedes, and which can re-spin shall be referred to as stage I larvae. Stage II 
larvae lack the ability to protract and retract their postpedes. At the outset of this 
stage they can locomote readily, but this capability is soon lost. Spinning activity 
also ceases during this period. Stage III insects cannot walk, nor can they right 
themselves. They twist their abdomens vigorously from side to side, an activity 
which aids in shedding the larval exuviae. Stage III includes the "pharate" 
("cloaked") pupal stage (Hinton, 1958). It terminates at ecdysis with the un- 
covering of the "white" pupa. 



NERVE CORD SHORTENING , 



3 5 

i "5 





CN 


- 


00 
<N 


O 
rN 





c- 


00 IN 


00 1 

q 


q 


9 


; 


o 


5 


q 


11 J 1 

+i 


1" 


?* 


o'_ 
-H- 


o' 

-H- 


o' 
-H- 


o o' ,^ 
-H-H-* 
oq 


!?i 


55-** 


5- 


_* 


=.* 


- r 
q* 


IF 


a *" 


K ~ I 


"* 


fO 


i-O 


?N* 


rs 


# 


^*" 


^>~ 


* 


*" 


* 


' 





g 














* 


* 


* 


* 


* 




* 


* 


* 








| 
























CO 


!2 


2 


00 





2 


o 


2 


00 


s 





o 


q 


q 


9 


qi 


'o 


fa 


52 


o'_ 
-H- 


5~ 


5 2 


o' 

-H 2 


~H - 


5- 


5 


52 






3 "' 


VO 


00 


a 


>o 


* 


' 


o 


o- 


00 


-o 


10 


>O 


IO 


>o 


1O 
















d 


d 


d 


d 


o' 


d 


d 


o" 


00 


q 


Id 








IN 




















h-T 

1 


fa 




52 


o' 
-H" 


fa 








1 


1 




1 


1 




o 


o\ 


00 


o 


o 


& 






















d 


d 


d 


d 


d 






















<M 


q 




o 


- 


00 

q 


00 


2 


2 





00 


2 


o 



^1 


2 


f 

10 


5? 


5^ 


52 


5~ 


^2 


5 2 


d 

-H - 


5 - 


5- 


^2 


52 


5=* 


il - 


rN 




* 


^ 


<"O 


[*3 


f*3 


~i 


CM 


CN 


"2 


"* 


"*. 


Tf 


q 


^* 




~ 


" 


"" 


*^ 


11 


^ 


"" 


"^ 


" 


1-1 


' H 


1-1 


CN 


"- 1 




"" 


- 


00 

q 


rs 


- 


00 

q 


2 


9 


00 

q 


2 


2 





o 





10 

1 

-t 


I s 


5^ 


52 


5- 


o' -. 
-f|2 


5 2 


d 

-H - 


r-l 


5= 


52 


^2 


^ 


o' ^ 
-H - 


1 5 




- 


-t 


- 


- 


"5 


"2 


CN 


CN 


- 


- 


"2 


CN 


o 


2 






Q 

































q 





c 





q 


O 


CO 

o 


o 


o 


q 


-t* 





-t 


= 


^ 


Hi- 


fa 


52 


c ^ 


Q 
_M ^ 


d 

-H - 


o' 
-H- 


5- 


^2 


0=0 


3 ^ 


d ,, 

4J - 




t 


t 


: 




-' 


- 


- 


-' 


CN 






- 


2 


CM 




c 


OO 


6 


o 


o 


d 


Z2 


C* 


2 


- 


00 


oo 
c 


2 


-r 


rN 


CM 


CN 


5- 


c- 


|2 


00-,^ 

ti-H ? 

00 "5 


,^ 


d O c 

,, 

0010 


O ^ 


1? 


|2 


i ? 


5 2 


5 - 













d 


d 


o' ' 


d' 


d ' 


' 


CM 


CN 


CM 


CN 


CN 














* 




# 


* 


-S 





*- 


* 


Jf 




O 


CN 


- 
























1 


5^ 


5^ 


5- 






! 




















- 


d 


d 





























00 




























q 































d 


o' ~ 





























-H"3 


-H" 


1 




1 


1 










1 


| 


1 




fc - 


-t 


-c 




























d 


d 




























o 


-t 


2 





q 







q 














- 


5? 


fa 


d _ 
-H- 


o'_ 
-H- 


4l2 


=> -, 


d 

H - 


5 - 


1 




00 


= = 


C f^l 


C "i 


~ 


* 


CN 


q 


00 


t^ 


>o 


ir. 


t 
















"~" 


^~ 


^" 


d 


o' 


d 


d 


d 
















- 


q 





o- 



r^ 

q 


q 


g 


g 


o 
q 


q 


q 


O 


S 


10 




5 


s^ 


52 


5- 


52 


?i ""' 


o 
-H - 


+1 - 


= - 


~^~ 


5s 


= = 


= ~l 


3 ~' 







q 


00 


t^. 


^c 


ID 


-t 


+ 


-t 


(*5 


f*5 


1*5 


-t- 


LT, 




~~ 


~ 


d 


d 


d 


d 


d 


d 


d 


d 


c; 


d 


d 


d 




10 

o 


IO 

o 


1O 

o 


s 


q 


>o 
q 


o 





o 


s 


1 


o 





5 


I 



o' 
-H 


II ^ 1 


5= 


5= 


52 


3 "^ 


d 


= ~i 


o'_ 


ii- 


is 


ii~ 




d 

-H - 


Jl 


' 





* 


^~ 





CN 


f^ 


fO 


10 


~t 


1O 


^ 


00 


X 




C 


d 


d 





d 











o' 


6 


d 


d 


d 


d 










i_ 


u 






l_ 










V_ ^ 


V_ ^ 


l 


|| 


rt 3 


^ i- 

b/j ^" 

rt 3 


QJ 

to-S 

r 


"" ^2 

si 


- 1 

r-4 3 


7 1 


o- J< 

3 i 


o| 


:| 


sl 


ll 


^ r"s 




w 


c/j 


W<5 




2 


z 


X 


~ X 


X 


^X 


"'z 


'Z 


"* x 


3- 3^ 


"=~x 




























< 


73 



:x 

_0 

c 
o 



M 

S 
J 



M 

rt 



2% 



RUDOLPH L. PI PA 



Stage of pupal development was defined by the number of hours after ecdysis. 
Pupal age-class intervals of five hours were used throughout (Table I). An in- 
termediate time for each interval is designated on the abscissas of Figures 2 and 3. 
Thus, individuals in the age class 0-5 hours are considered 3 hours old ; those in 
the age class 6-11 hours, 9 hours old; etc. Stage II precedes ecdysis by 19-27 
hours (25 of 33 cases recorded) ; stage III by 6-15 hours (18 of 23 cases recorded). 




IMCUKK 1. Dorsal view of central nervous systems of Gallcria. Fixed in situ with 95% 
ethanol, removed, and preserved in 10% formalin. (A) Stage I larva; (B) Pupa, 12-15 hours 
after ecdysis; (C) Adult female. AB1, AB2, AB3, AB6, AB7.8, Abdominal ganglia; SG, 
Suhesophageal ganglion; TH2, TH3, Thoracic ganglia; VD, Ventral diaphragm. 






NERVE CORD SHORTENING 



297 



100 GA 



90 



80 



70 



*0 



A=I-H 

o=n-in 



= 6+7,8 



5 



O 

Z 
UJ 



60 



50 



5 



^40 




i Hi 



o o 

O 

* * 



* ( 
* 



.. A 
O 

O 



30 



20 



10 



-25 -15 -5*5 Ib 23 35 $ 55 -25 -15 -5*5 15 25 35 45 55 

HOURS ADULTS 

FIGURE 2. Per cent of mean initial (stage I) lengths remaining at progressive develop- 
mental stages. Standard deviations indicated by vertical lines, except where obscured by a 
symbol. (A) I-II, connectives between first and second thoracic ganglia; Il-III, connectives 
between the second and third thoracic ganglia; (B) 6 + 7,8, length from anterior border of 
abdominal ganglion six to posterior border of abdominal ganglion eight ; II + III + 1 + 2, length 
from anterior border of mesothoracic ganglion to the posterior border of the second abdominal 
ganglion. P, time of ecdysis. Percentages of stage I lengths remaining in adult males and 
females included at the extreme right. 

Accordingly, data for stage II insects are plotted at 23 hours; those for stage 
III at 11 hours. Data for stage I are plotted at 26 hours, regardless of the fact 
that many of these insects were probably further removed from ecdysis than this. 

I 'ATTKKN OK REORGANIZATION 

The ventral nerve cord of a stage 1 larva (Fig. 1A) consists of twelve de- 
finitive ganglia, all fairly uniform in size. The first eleven are associated by 



298 



RUDOLPH L. FIFA 



0.75 



0.50 



0.25 



-25 



0=6+7,8 

o=3r-ur 
A=I-H 




-15 



-5 i 5 15 

HOURS 







55 



o c 

ADULTS 



FIGURE 3. Mean lengths of selected segments of the ventral nerve cord during progressive 
developmental stages. Values for adult males and females are included at the extreme right. 
Standard deviations indicated by vertical lines. Symbols as in Figure 2. 



paired connectives which vary in length. These are visible at low magnifications. 
The seventh and eighth abdominal ganglia, however, are contiguous. The ventral 
nerve cord is associated with the brain by a pair of circumesophageal connectives. 
Comparison with the ventral nerve cord of a newly emerged adult (Fig. 1C) 
reveals prominent differences. The adult possesses nine externally recognizable 
ventral ganglia ; the first, second, and sixth abdominal ganglia of the larva have 
vanished. The paired nature of the connectives between the subesophageal and 
first thoracic ganglia and between abdominal ganglia is no longer distinct. The 
second and third thoracic ganglia, lacking intervening connectives, are now con- 
tiguous. The connectives between the subesophageal ganglion and brain, and be- 
tween the first and second thoracic ganglia, have also shortened ; those between 
the subesophageal and first thoracic ganglia have lengthened. There is a striking 
disparity in comparative volumes of thoracic and abdominal ganglia, a reflection of 
dishannonic growth along the cranio-caudal axis. Another difference is that the 
adult abdominal interganglionic connectives are attached to the ventral diaphragm 
(VD, Fig. 1C), which, in turn, inserts on the abdominal exoskeleton. Contrac- 



NERVE CORD SHORTENING 299 

tions of muscle fibers in the diaphragm cause the abdominal nerve cord to lash 
from side to side within the hemocoele. 

Sequential measurements of the connectives located between the first and sec- 
ond (I-II) and second and third (II-III) thoracic ganglia indicate that they lose 
35-45% of their mean initial lengths before ecdysis (Fig. 2A). By 30-40 hours 
after ecdysis connectives I-II shorten maximally, losing about 75% of their origi- 
nal length. This contrasts with connectives II-III which continue to diminish 
until ca. 45 hours, when they disappear, leaving the ganglia contiguous. 

It might be suspected that decrease in the length of I-II is not due to shortening, 
but is simply a result of overgrowth by the first and second thoracic ganglia. Such 
an interpretation is not supported by comparative measurements. These show that 
no more than half the 0.6-0.7 mm. lost could possibly be accounted for in this 
manner, and then only if all growth were directed toward I-II. 

Shortening of I-II and lengthening of connectives between the subesophageal 
and first thoracic ganglia (SE-I) are not in phase. By 12-17 hours after ecdysis, 
when elongation of SE-I is first detectable, shortening of I-II is approximately 
75% complete (Table I). Partial temporal separation of these two morphogenic 
processes occurs elsewhere within the ventral nerve cord, and shall be mentioned 
again below. 

Reorganization, which results in disappearance of the first, second, and sixth 
abdominal ganglia, is indicated by Figure IB. The first and second abdominal 
ganglia approach the third thoracic anteriorly, while the sixth meets the contiguous 
seventh and eighth posteriorly. During shortening the intervening connectives in- 
crease in diameter until the}- are nearly as \vide as the ganglia which are about to 
coalesce. The discreteness of the connectives involved is lost before shortening is 
concluded. This has necessitated measuring from the anterior border of one 
ganglion to the posterior border of another. Shortening which results in concen- 
tration of the first two abdominal and the second and third thoracic ganglia was 
approximated by measuring from the anterior border of the second thoracic 
ganglion to the posterior border of the second abdominal ganglion (II + III + 
1+2, Table I; Fig. 2B). Shortening, which results in coalescence of abdominal 
ganglia six, seven, and eight, was followed by measuring from the anterior border 
of six to the posterior border of eight (6 + 7,8 Table I ; Figs. 2B ; 3). 

The entire second abdominal ganglion does not coalesce with the fused meta- 
thoracic-first abdominal complex. Instead, anterior migration of the definitive sec- 
ond abdominal ganglion ceases 12-17 hours following ecdysis (Fig. IB). After 
that time an opaque white mass, presumably consisting of internal components of 
the second abdominal ganglion, continues forward. As this proceeds the second 
abdominal ganglion diminishes until an inconspicuous "hull" remains. The latter 
subsequently moves back as connectives between it and the advancing white mass 
elongate. The "hull" disappears 30-35 hours after ecdysis. Shortening, as de- 
termined by measuring II + TIT + 1 + 2 (Table I), is about 70% complete 12-17 
hours after ecdysis, when elongation of collectives between the "hull" and white 
mass first becomes apparent. 

Abdominal interganglionic connectives 23. 3-4, 4-5, and 5-6 do not shorten 
appreciably. Application of the /-test for significance between the highest and low- 
est means obtained for each during metamorphosis yielded t- values of 0.08, 0.16. 
0.11, and 0.10, respectively. 



300 RUDOLPH L. PIPA 

Events resulting in coalescence of the sixth abdominal ganglion with seven and 
eight resemble those noted during fusion of the first two abdominal ganglia with the 
third thoracic. Thus, it appears that the contents, but not the "hull" of ganglion 6 
become incorporated with 7,8. As connectives 6-7,8 shorten they increase iti 
diameter, and it becomes increasingly difficult to separate boundaries of the 
ganglia (Fig. IB). Consecutive measurements of 6 + 7,8 reveal that coalescence 
is completed 30-40 hours after ecdvsis (Figs. 2B, 3). 

When the mean lengths of selected segments of the ventral nerve cord are 
plotted against developmental time, the distributions depicted in Figure 3 are 
obtained. The origins and slopes of three of these were estimated by assuming 
linearity, and by applying the method of least squares. Extent of shortening per 
24 hours approximated in this manner is 0.27 mm. for connectives I-II (Y : 1.024 
+ 0.0112X); 0.46 mm. for II-III (Y : 1.37 + 0.0192X) ; and 0.41 mm. for 
6 + 7,8 (Y = 1.73 + 0.0171X). Tests for homogeneity of regression on each of 
the three combinations of the three lines were significant in each case (I-II vs. II- 
III, t =: 11.56***; I-II vs. 6 + 7,8, t == 9.96*** ; and II-III vs. 6 + 7.S. / == 2.38* ). 

DISCUSSION 

The consecutive measurements made during the course of this study clearly 
demonstrate that ganglionic concentration is accomplished by shortening of in- 
tervening connectives. The explanation proposed by Murray and Tiegs (1935) 
for ganglionic concentration in the beetle, Calandra orysac, namely, that it is due 
to a "proliferation of cells" and subsequent overgrowth, cannot be accepted for 
Galleria. The present study does not indicate which of the nerve cord com- 
ponents is responsible for decrease in length of the connectives, but the conclu- 
sion that it is caused by extraneuronal cellular migration, neuron shortening, or 
both, seems unavoidable. 

The extent to which different sectors of the ventral nerve cord shorten is vari- 
able. The connectives between the second and third, third and fourth, fourth and 
fifth, and fifth and sixth abdominal ganglia do not shorten significantly. Con- 
nectives between the first and second thoracic ganglia shorten approximately 75% 
of their mean initial length, while those between the second and third disappear. 
Connectives between the third thoracic and first abdominal ganglia, first abdominal 
and second abdominal ganglia, and abdominal ganglia 6 and 7,8 not only disappear, 
but their associated ganglia coalesce. Thus, the following gradations prevail : ( 1 ) 
No significant shortening. (2) Partial shortening. (3) Complete shortening with 
establishment of contiguity between centers. (4) Complete shortening with 
coalescence of centers. 

Gross features predict extensive concomitant modifications at the tissue and 
cellular level during shortening. The composition and fate of the ganglionic con- 
stituents of the advancing white mass and the shortening connectives proper 
are but a few of the manifestations which require histological and cytological 
clarification. 

Ventral nerve cord shortening, well under way before ecdysis, is concluded 
30-45 hours after. The gross features of neurometamorphosis described here re- 
quire approximately three days for completion, or approximately 33% the mean 
total time from onset of pupation (stage IT) to adult emergence. 



NERVE CORD SHORTENING 

Much of the shortening which occurs is compensated by subsequent elongation. 
Extent of shortening of connectives I-II and extent of elongation of connectives 
between the subesophageal and first thoracic ganglia are nearly identical (Table I). 
Little, if any, decrease in total ventral nerve cord length results. Similarly, loss 
of interganglionic connectives between the third thoracic and first abdominal 
ganglia, and between the first and second abdominal ganglia is entirely compensated 
by elongation between the second and third abdominal ganglia. 

The resultant effect is an adult ventral nerve cord only slightly shorter than 
that of the stage I larva from which it has developed. Of greater significance, 
perhaps, is the correlation between the gross structural reorganization of the 
ventral nerve cord and the skeleto-muscular system. During metamorphosis the 
thoracic and first two abdominal metameres are brought closer together, the pro- 
thoracic, metathoracic, and first two abdominal segments suffering extensive re- 
duction during the process. Shortening of the corresponding interganglionic 
connectives is in accord with these changes. 

Statistical analysis of these data is principally due to the generous efforts of 
Dr. Howell V. Daly. I thank Dr. Roderick Craig and Dr. Howell V. Daly for 
critically reading the manuscript, and for relevant suggestions. The technical 
assistance of Mrs. Nancy Luykx is also gratefully acknowledged. 

SUMMARY 

1. Gross features of interganglionic connective shortening during metamorpho- 
sis of Gallcria mcllonella (L.) are described. Gradations range from no significant 
shortening, to partial shortening, to shortening with establishment of contiguity 
between ganglia, to complete shortening with coalescence of ganglia. 

2. Under the experimental conditions employed, shortening commences about 
a day prior to ecdysis, and is completed 30-45 hours after ecdysis. The rates at 
which various connectives shorten differ significantly from one another. If 
linearity is assumed, these range from 0.3 to 0.5 mm. per day. 

3. Much of the shortening is compensated by subsequent elongation of con- 
nectives. The two morphogenic processes are not in phase; shortening is 70-75% 
complete before elongation can be detected. 

4. The adult ventral nerve cord is about 1520% shorter than that of the stage 
I larva from which it has developed. Shortening has altered the relative locations 
of certain of the ganglia so that they are in accord with structural reorganization 
of the skeleto-muscular system. Not only are ganglia retained close to their 
effector organs as a consequence, but conduction times between certain of the 
centers would be expected to be reduced. 

LITERATURE CITED 

BRANDT, E., 1879. Vergleichend-anatomische Untersuchungen iiher das Nervensystem der 

Lepidopteren. Horae Soc. Entoinol. Ross., 15: 1-16. 
BUXTON, P. A., 1931. The measurement and control of atmospheric humidity in entomological 

problems. Bull. Eut. Res., 22: 431-457. 
GILBERT, L. L, AND H. A. SCHNEIDERMAN, 1961. The content of juvenile hormone and lipid in 

Lepidoptera: sexual differences and developmental changes, (icn. Coinp. Endocrin 1- 

453-472. 



302 RUDOLPH L. PIPA 

HINTON, H. E., 1958. Concealed phases in the metamorphosis of insects. Sci. Prog., 182: 

260-275. 
MURRAY, F. V., AND O. W. TIEGS, 1935. The metamorphosis of Calandra or\sac. Quart. J. 

Micr.Sci., 77: 405-495. 
NEWPORT, G., 1832. On the nervous system of the Sphinx ligustri, Linn., and on the changes 

which it undergoes during a part of the metamorphosis of the insect. Phil. Trans. Rov. 

Soc. London, 122: 383-398. 
XKWPORT, G., 1834. On the nervous system of the Sphinx ligustri, Linn., (Part II) during the 

latter stages of its pupa and its imago state ; and on the means by which its development 

is effected. Phil. Trans. Roy. Soc. London, 124: 389-423. 
PIEPHO, H., 1950. Hormonale Grundlagen der Spinntatigkeit bei Schmetterlingsraupen. 

Zcitschr. TicrpsychoL, 7: 424-434. 
WIEDBRAUCK, J., 1955. Vom Spinncn bei Schmetterlingsraupen und seiner Abhangigkeit von 

Metamorphosehormonen. Zcitschr. TicrpsychoL, 12: 176-202. 
WILLIAMS, C. M., 1961. The juvenile hormone. II. Its role in the endocrine control of 

molting, pupation, and adult development in the Cecropia silkworm. Biol. Bull., 121: 

572-585. 
YEAGER, J. F., 1939. Electrical stimulation of isolated heart preparations from Pcriplanchi 

amcricana. J. Agr. Res., 59: 121-137. 



X-IRRADIATION-INDUCED CONGENITAL ANOMALIES IX 

HYBRID MICE 1 

ROBERTS RUGH AND MARLIS WOHLFROMM 

Radiological Research Laboratory, Department of Radiology, College of Physicians & Sun/eons, 

Columbia University, New York 32, N. Y. 

Heterosis, or hybrid vigor, in radiobiological studies, has been demonstrated 
for the adult mouse (Rugh and Wolff, 1958) ; for the testes (Rugh, Funk and 
Wohlfromm, 1961) and for some induced congenital anomalies (Rugh, Wohlfromm 
and Grupp, 1961). It was further shown (Grahn, 1958) that there are non-addi- 
tive (heterotic) effects when a sensitive strain is crossed with a particularly re- 
sistant strain (C57BL/6 X BALB/c), while the variance in sensitivity, as meas- 
ured by the dosage-mortality slope, appears to be merely additive. It was decided 
to carry the experiment one further generation, by crossing the hybrids with them- 
selves as well as with each of the parental strains, in order to detect, if possible, any 
influence caused by the preponderance of either genotype. The test was made 
by x-irradiating the embryos at 8.5 days' gestation and determining the effect at 
18.5 days, just prior to the expected delivery. 

MATERIALS AND METHOD 

The mice used were the CF1 Swiss white strain from Carworth Farms and 
the C57 BL/6 Blacks from the Bar Harbor Laboratory. These strains are easily 
interbred and give viable offspring. All matings were made overnight and the fe- 
males with vaginal plugs were separated the following morning and marked as 0.5 
day pregnant. While this exposure of the females could give a range of conception 
of 16 hours in simultaneously exposed mice it is now known that most of the 
matings occur early in this period and that all of the mice were at least 0.5 day 
pregnant as of 9 A.M. the morning following the introduction of males. Since 
the x-irradiations were to occur at 8.5 days it is believed that the time range of 
conception was somewhat less significant than if the exposure were shortly after 
conception. The number of pregnancies for any combination was at least 21 so that 
the time variance in fertilization is averaged out. 

Whole body x-irradiation was achieved with parallel tubes in cross-fire, each 
at 67 cm. distance from the gravid uterus (filtration 0.28 Cu and 0.50 Al, half 
value layer 0.6 Cu, dose rate 50 r/min., total dose 200 r). The machine, was run 
at 184 KVP, and 30 MA. The mice were not anesthetized, but were placed in a 
plastic cage within which the dosimetric determinations were made. 

Since mice often destroy or eat their offspring, when they are abnormal, it 
was necessary in this study to sacrifice the pregnant mice at 18.5 days (prior 

1 Based in part on work performed under Contract AT (30-1) -2740 for the U. S. Atomic 
Energy Commission and aided by Grant RH 97 from Div. Radiological Health, Bur. State 
Services, Public Health Service. 

303 



ROBERTS RUGH AND MARLIS WOHLFROMM 

to the expected delivery) and analyze the effects of x-irracliation at that time. 
It was also found that there was considerable variation between and within 
litters, even among the controls, so that to provide adequate statistical data, a 
minimum of 21 (and maximum of 42) pregnancies per set of categories was 
collected. The minimum number of implantations lor any set of categories was 235 
(and the maximum was 471), either figure giving a statistically adequate number 
so that anomaly percentages have significance. Each pregnant mouse was killed 
by cervical dislocation and the gravid uterus dissected out immediately. On the 
basis of prior studies, as well as new current findings, the categories included 
among the anomalous conditions were resorptions, dead fetuses, and obvious CNS, 
eye, visceral and tail anomalies. The "normals" included all which appeared 
grossly normal, but all or many of these could exhibit microscopic anomalies not 
readily apparent. It cannot be assumed that a "normal-appearing" fetus in a litter 
where there are gross anomalies caused by irradiation will itself be entirely normal. 
Thus, this paper categorizes only those gross anomalies which are readily identi- 
fiable, and classes as "normal" all fetuses which appear grossly to be normal. 

EXPERIMENTAL DATA 

The data are presented herein by means of tables which include all of the im- 
plantations examined, namely some 296 pregnancies and 2780 implantations. It 
can be seen readily that 200 r at 8.5 days affects the C57 embryo much more 
drastically than it does the CF1 embryo of the same age, reducing the "normals" 
to 0.4% and 40%, respectively. It had previously been shown that the first 
generation hybrids were more resistant to this x-irradiation insult than were 
either of these parent stocks (Haverland and Gowen, 1960; Rugh, Wohlfromm 
and Grupp, 1961; Nash and Gowen, 1962). However, when one obtains the F 
by mating the first hybrid generation, it is found again that there is a higher 
radioresistance in the appearance of more "normals" (53%). This may be sug- 
gestive of heterosis. When the first generation hybrids are crossed with either 
of the original parental stocks, it is found that slight heterosis is still indicated 
when the parent genotype is CF1 but not when it is C57. However, even in this 
latter cross the results are more favorable than in the C57 X C57 cross, simply 
because of the presence of CF1 genes, and there are as many normals as there are 
in the control cross of CF1 X CF1. This means simply that either the influence 
of the CF1 genes, in preponderance, affords some radioresistance to the embryos 
or when the C57 genes are in preponderance the initial first generation heterosis 
is reduced (see comparable data of Nash and Gowen, 1962). However, in every 
cross tried, the original C57 genotype, enforced by any CF1 genes, proved to be 
at least as radioresistant as the F 1 of the CF1 X CF1 cross. This suggests a 
radioresistant influence comparable to dominance. 

It is known that more eggs are ovulated than are fertilized, a fact based upon 
the number of corpora lutea and the number of implantations (Otis, 1953). The 
data of this paper begin with implantations. Among these are some that are 
destined to die and be resorbed, in both strains of mice. But since the percentage 
of resorptions in the control CF1 and the control C57 mice were about the same 
proportion, they are included in all calculations. 

The so-called "normal" mice may not, in fact, be completely normal. They 



CONGENITAL ANOMALIES IN HYBRID MICE 



305 



"9 
-5 







w S 

-J s 

PO 53 
**! 

H 
_*- 

8 



S 

O 



g 



CO 






~ 








nj 






' 


NC 


1 - 




jj 






'-i 


0| 


f- 




O 





C 


OO 


(^"i J^.. 


O ^ i 




c/5 






01 









c 






NO 


, , 


. , 




.2 


PC CM 


CM 0) 


OO 


^ Ol 


'"". OC 




4-1 

a 


PC 


O-. 


L-.' 


OO CM 






IH 

o 














t 


O NO 


^- 


C NO 


01 00 


r- 




<u 


NO *^* 


CM 


CM -t- 


01 LO 


01 ON 




i - 




PC -f 


00 




NO 




Uj ._ 












-H 




w 





-H - 


01 





c -t- 




tj 


LO 


g 












OO 


PC f 


CM 


' 1 


OO 




O) 





CM' o 


LO 


-t-' 


04 




i> 


oi 


NO ^ 


C' PC 


C c 


r- 




M 






"~ 


t-H 






U 

o u 


t- -t 


CM^ 


r . 


^ 


-t i^ 




" ** 


' C 





O PC 


NO' 


-t-' 




5 


- 


' 


*_- ^-s 


- 


- * -^ ' 




1" 


OO 


PC cO 


01 0- 


C NO 


~ ~ 




CO 

_CLJ 






^ 


^ 








~? 


CM 


rc 00 


PC NO 


^ So 




2 E 


^ 


< 


C CM' 


O 1~~' 







c 


c PC 


PC c 


i r^. 


c 


^-H CM 




ai 
















10 








^~. 00 




c 


O CM^ 


PC ^H 


NO" 


"1 ? 


f 




s 


i 


ci i 












G 


f ON 


NO OO 


C ON 


PC 1~*" 









CM 


~* 


PC 


PC 


fC 




^ 


>o o" 


o . 


PC" 


O~ NO" 


~ c 







OO -f 


oo o 


ON LO 


ON "f 


ON -t- 




u 
C 


ON "^ 


PC 


NO CM 


t^ Ol 


OO 01 




Z 


ON ON 


C5 


-T PC 


Tf CM 


PC O 




- 


PC 


CM 


CM *-c 


CM -H 


O] *-^ 




4-1 


-H LO 


PC CM 
LO LO 


ON 00 
NO -f 


PC -t 
t^ 1 " \O 


NO LO 


1 


o 


-f CN 


fM 01 


01 01 


0] PM 


Ol 01 


01 


^ 

< 














CO 




PC 'f 


OO OO 


OO r^ 


OO t^ 


NO 


?^ r^ 


-f fM 


PC PC 


O] Ol 


0] Ol 


01 0) 


ON 


Cl- 03 












01 


b 
















U X 

\/ 


U X 
\/ 


U X 

\/ 


U X 

\/ 


U X 

\/ 






II 


II 

10 


II 


II 


II 

'O 




CO 
CO 



o 


u 

X 


U 
X 


u 


u 

X 


u 

X 




01 


t 


t~~ 


jy-, 


, 


-H 




CO 
3 




ta 
U 


10 

U 


U 
X 


PL, 
U 


J- 

u 

--. 










7? 


LO 


LO 


J/> 








U 


u 


u 


cd 








^^ 






O 








LO 






H 








U 









f> 

0) 



ty 
I 



en 

<u 



O 

o 
u 
QJ 



bo 
f. 
' 



n) 

u 



CM 

O 



5'T 

ux 



306 



ROBERTS RUGII AND MARLIS WOHLFROMM 



were normal-appearing mice since they were the survivors of x-irradiatiofis which 
conditioned anomalies in the same litter, hut it is yery doubtful that they themselves 
were normal. The "dead" were those which suryiyed until the third trimester 
and had the distinguishable features of fetuses. Those which were anemic and/or 
stunted may have been moribund, but were alive at the time of analysis. The eye 
defects, normally seen in C57 stock, were expressed as microphthalmia, anoph- 
thalmia, etc. In a few cases evisceration was found, in which the developing 
visceral organs failed to be enclosed in oyergrow r th of the abdomen. 

Among the anomalous conditions it is important to point out that the controls 
are not without congenital anomalies which may well lie of genetic origin ( Ebert, 
1961). Howeyer, the control data must be taken as the base line for comparison, 
in order to determine the additive effect of x-irradiation insult. Such added con- 
genital anomalies are not genetic, but developmental. For instance, resorptions 
(which generally mean early embryonic death) are about the same in the two 

TABLE II 
Percentage dead and anomaly risk 



Group 


# Implanta- 
tions 


# Dead & 
resorbed 


# Survivors for 
CNS anomalies 


# & % of CNS 
anomalies 


CF1 X CF1 controls 
CF1 X CF1 + x-rays 


471 
235 


64 (13.6%) 
105 (44.7%) 


407 
130 


(0%) 
33 (25.4%) 


C57 X C57 controls 
C57 X C57 + x-rays 


253 
252 


37 (14.6%) 
249 (98.8%) 


216 

3 


3 (1.3%) 
(0%) 


C57/CF1 X C57/CF1 controls 
C57/CF1 X C57/CF1 + x-rays 


269 

248 


20 (7.4%) 
85 (34.3%) 


249 
163 


1 (0.4%) 
7 (4.3%) 


C57/CF1 X CF1 controls 
C57/CF1 X CF1 + x-rays 


273 
264 


25 (9.2%) 
95 (36.0%) 


248 
169 


1 (0.4%) 
20 (11.8%) 


C57/CF1 X C57 controls 
C57/CF1 X C57 + x-rays 


261 
254 


21 (8.0%) 
127 (50%) 


240 
127 


1 (0.4%) 
2 (1.6%) 



control stocks (12% and 13 % ) but are lower in the controls of all hybrid crosses. 
This would constitute added evidence of heterosis. When the embryo is x-irradi- 
ated at 8.5 days the resorption data range from 18.6% to 92%, indicating great 
variation in response. The highest level of resorptions occurs among the pure 
C57 lines (92%), indicating maximum embryonic radiosensitivity, and is lowest 
among the F x generation of the hybrid X hybrid cross. In the previous study 
(Rugh, Grupp and Wohlfromm, 1961) it was shown to be 90% for the C57 line. 
When the hybrid generation is crossed to the C57 stock, the resorptions are again 
high (38%) and lowest when the hybrid is crossed with the hybrid. Thus, as 
even among the control data, it appears that there is radioresistant heterosis w-hen 
CF1 and C57 are crossed, and this heterosis is modified by outcrossings in such a 
way as to indicate that the presence of any CF1 genes is beneficial. When hetero- 
zygosity appears to be maximum, the conditions for surviving the radiation insult 
are the greatest (see Haverland and Gowen, 1960). 



CONGENITAL ANOMALIES IN HYBRID MICE 



307 



Consolidating the deaths, whether late fetal or early embryonic death, with 
resorption and considering onlv the survivors as having anv risk of developing 
anomalies, it is obvious that the percentage of congenital anomalies of the central 
nervous system was always greatest when the IT 1 collective influence was the 
greater. The resorption percentages were greatest when the C57 influences were 
the greater. This sort of experiment does not reveal any specific genetic influences, 
hut rather lumps together the CF1 influences and compares them with the ag- 
gregate effect of C57 influences on the development. The anomalies are no doubt 
developmental, without mutational factors involved, because they arise from x-ir- 




PLATE I 



308 



ROBERTS RUGH AND MARLIS WOHLFROMM 




CONTROL 




MICROCEPHALY 

Right side 4 




MICROCEPHALY 

Left side 5 




MICROCEPHALY 

AND 
EXENCEPHALY 




GROSS CEPHALIC 
ANOMALY 



X- IRRADIATION- INDUCED 
CENTRAL NERVOUS SYSTEM ANOMALIES 

PLATE II (drawings by P. Van Dyke) 



CONGENITAL ANOMALIES IX HYI5KII) MICK 

radiation at almost mid-gestation. However, having indicated a genetic prepotency 
toward early death and resorption, in the case of the C57 mice, and toward CNS 
anomalies in the CF1 mice, it is necessary to suggest that this difference may he 
due to the fact that the hasic genotype of the CF1 mice may be hardier (with 
regard to radiosensitivity) so that when its influences are preponderant, there is 
a better opportunity for CNS congenital anomalies to develop simply because 
there is better survival of those with CF1 genetic factors. 

Among the specific anomalies listed, those involving the central nervous sys- 
tem appear to be the most obvious and drastic, probably since 8.5 days represents 
the time of initial and most active neurogenesis. (See Plates I and II.) However, 
interesting data appear here in that the incidence of these congenital anomalies is 
more frequent among the progeny of CF1 mice than among the supposedly more 
radiosusceptible C57 strain. This is explained in part by the fact that the C57 
strain is so radiosensitive that most of them are lost as resorptions. Among 
those lost were probably many that, had they survived, might have exhibited CNS 
anomalies. None of the few (1%) surviving x-irradiated C57 mice had gross CNS 
anomalies. This difference is further borne out by the fact that when the hybrid 
mice (CF1 X C57) are crossed with the CF1 strain more (7.6%) CNS anomalies 
develop (because of better survival) than when such hybrids are crossed with 
the C57 strain (0.8%). All x-irradiations were at 8.5 days' gestation. It is 
suggested that differences following the back-cross to the two strains may be re- 
lated to the higher incidence of resorptions whenever the C57 genes are present, 
thus reducing the fetuses among which CNS anomalies could occur. Other de- 
velopmental anomalies, such as stunting, anemia, evisceration, and shortened tails, 
did occur but not in sufficient numbers for evaluation, although the data are sug- 
gestive. These conditions cannot be of genetic origin, since the x-irradiations oc- 
curred after 8.5 days of embryonic development. They are achieved by x-irradiation 
interruption of the normal morphogenetic processes, aided by the subsequent deletion 
of necrotized 'blast cells of the actively differentiating embryo. Since eye anomalies 
normally occur in pure C57 stocks one would expect the C57 genie influences to 
be correlated with higher incidence of these defects, but this is not borne out by the 
data. Again this may be due to the obscuring of this tendency by the presence of 
CF1 genes which are not predisposed to eye defects. The question may be asked 
as to whether it might be better to be so radiosensitive that there is a high death 
rate rather tlian to be less radiosensitive and survive, only to develop congenital 
anomalies. 

SUMMARY AND CONCLUSIONS 

1. Increased radioresistance is manifested in hybrid embryos exposed to 200 r 
at 8.5 days of embryonic development when divergent stocks of mice are cross- 
bred, as determined by the incidence of congenital anomalies. This is evidence of 
embryonic heterosis. 

2. When hybrid mice are mated with hybrids (of the same crosses) or with 
either of the parental stocks, it is apparent that the presence of CF1 influence 
(genes?) affords the embryos more radioresistance only when the cross is with 
the CF1 line. The presence of CF1 genes brings the results up to the pure CF1 
data so that it appears that there is a sort of dominance effect which becomes 
heterosis when the proportion of CF1 genes exceeds this minimum in these 
heterozygous embryos. 



310 ROBERTS RUGH AND MARI.IS WOHLFROMM 

3. In hybrid combiii:iti<ms the development of CNS congenital anomalies ap- 
pears to be more frequently related to the presence of CF1 genes than to those 
from the C57 strain. Conversely, resorptions appear to be more directly related to 
the presence of the C57 genes. Even in the pure strains this is substantiated be- 
cause 14% of the CF1 embryos and none of the C57 embryos showed CNS 
anomalies following 200 r x-rays at 8.5 days' gestation. Likewise, the C57 strain 
embryos reacted to this exposure by producing 92 % resorptions while the CF1 
embryos showed only 32%. 

4. The data on congenital anomalies seem at first to be confused. This is clue 
to the simple fact of relative radioresistance of the CF1 embryos, which allows 
them to survive and hence to develop x-irradiation-induced congenital anomalies, 
particularly of the central nervous system. The radiosensitivity of the C57 strain 
reduced their survival. As a result of this, there are fewer survivors to develop 
CNS anomalies. 

5. As in other studies in heterosis, it appears that this embryonic hybrid vigor 
is correlated with maximum heterozygosity and is reduced as this condition is 
diluted toward either of the parental conditions. 

6. Radioresistance (or, conversely, radiosensitivity) of the embryo appears to 
be closely allied to inherent genetic factors quite different in the two strains of 
mice. In one strain there is the greater tendency to resorption and death and in the 
other to survival with attendant development of congenital anomalies. 

LITERATURE CITED 

EBERT, J. D., 1961. First International Conference in Congenital Malformations. /. Chron. Dis., 
13: 91-132. 

GRAHN, D., 1958. Acute radiation response of mice from a cross between radiosensitive and 
radioresistant strains. Genetics, 43: 835-843. 

HAVERLAND, L. H., AND J. W. GOWEN, 1960. Physiological vigor as a factor in radiation reduc- 
tion of lifetime fertility in mice. Rad. Res., 13: 356-368. 

NASH, D. J., AND J. W. GOWEN, 1962. Effects of x-irradiation upon postnatal growth in the 
mouse. Biol. Bull., 122: 115-136. 

OTIS, E., 1953. Prenatal mortality rates of seventeen radiation-induced translocations in mice. 
Univ. Rochester A.E.C. Pub. #291. 

RUGH, R., AND J. WOLFF, 1958. Increased radioresistance through heterosis. Science. 127: 144- 
145. 

RUGH, R., H. FUNK AND M. WOHLFROMM, 1961. Heterosis and testes x-irradiation. Atom- 
praxis, 7 : 243-248. 

RUGH, R., M. WOHLFROMM AND E. GRUPP, 1961. Evidence of prenatal heterosis relating to 
x-ray induced congenital anomalies. Proc. Soc. Exp. Biol. Med., 106: 219-221. 




SUN COMPASS ORIENTATION OF PIGEONS UPON EQ 
AND TRANS-EQUATORIAL DISPLACEMENT 

KLAUS SCHMIDT-KOENIG 

Dcpt. of Zoology, Duke University, Durham, N. C., and Max-Planck-Institut f. 
Verhaltensphysiologie, Abt. Mittelstaedt, Wilhelmshavcn and Seewiesen, Germany 

The operation of the sun compass in northern latitudes has been established in 
numerous arthropods, reptiles, fish and birds (von Frisch, 1950; Pardi and Papi, 
1952; Birukow, 1960, Gould, 1957; Fischer, 1961; Hasler ct al., 1959; Braemer, 
1959; Kramer, 1952a, 1952b; von St. Paul, 1953). The utilization of this mecha- 
nism for actual field orientation has been demonstrated in many arthropods (von 
Frisch, 1950; Pardi and Papi, 1952 and others) and pigeons ( Schmidt-Koenig, 
1960, 1961) ; it has been shown likely that it is used by turtles (Gould, 1957) and 
fish (Hasler et al,, 1960). 

A large number of field studies of migratory birds has revealed the prevalent 
role of directional orientation (Riippel. 1944; Rowan, 1946; Riippel and Schiiz, 
1948; Perdeck, 1958; Lack, 1959, 1960). The actual application of the sun 
compass during migration has not yet been proven. However, if the sun compass 
were applied, birds whose migrational routes cover many degrees of latitude or 
lead to equatorial or trans-equatorial regions would encounter rather striking 
changes in the sun's angular velocity (both azimuthal and directional). Braemer 
(1960) particularly has pointed to the complex nature of a full scale sun compass 
which must operate at grossly different latitudes and different seasons. Some- 
thing comparable to an almanac would be necessary to master all significant 
variables. 

It is evident that an examination of the performance of the sun compass under 
extreme solar conditions is inevitable. Southward latitudinal displacement has 
been carried out in the past with bees by Kalmus (1956) and Lindauer (1959) 
and with fish by Hasler and Schwassmann (1960). 

In this, the first of a series of experiments, the present author examined the 
operation of the sun compass in a stationary training apparatus in homing pigeons 
raised at Durham, N. C., upon displacement to Belem, Brazil and Montevideo, 
Uruguay, in fall, 1961. We are aware of the fact that pigeons as non-migratory 
birds may well differ in their orientational capacities from migratory birds. This 
species was, nevertheless, chosen as experimental subject because it handles best 
in experiments and because the translocation expedition was expected to be too 
difficult to be accomplished on the first try with small migratory passerines. We 
will call upon those later. 



This work was supported by grants from the Xational Science Foundation 
(grants G-9816 and G-1984') to P. II. Klopfer and K. Schmidt-Koenig) and bv 
contracts from the Office of Naval Research (contracts 301-244 and 301-618 with 

311 



312 KLAUS SCHMIDT-KOENIG 

Duke University). I would like to particularly acknowledge the special efforts 
undertaken by both agencies to facilitate the expedition to South America. I am 
much indebted to Dr. H. O. Schwassmann for his stimulating and helpful role. 
The experiments reported here were initiated by discussions with H. O. Schwass- 
mann at the 25th Cold Spring Harbor Symposium in 1960. H. O. Schwassmann 
also must be credited for his effort in the founding of a colony of pigeons at Belem. 
Brazil, of birds from the Wilhelmshaven and the Duke University strain * which 
will be used for further joint experiments there. I am grateful to Dr. Peter H. 
Klopfer for his steady helpfulness and support ; further, to the late Dr. Walter A. 
Egler, Director of the Museu E. Goeldi, Belem, Brazil, and his successor Dr. 
Eduardo Galvao, to Dr. Jose Maria Conduru, director of the Institute Agronomica 
do Norte, Belem, Brazil, for their extremely helpful cooperation ; to Drs. Paul 
Ledoux and Werner Sattler (then at Belem) for their kind assistance; last, but not 
least, to Prof. Dr. Victor Bertullo, director of the Departamento de Investigaciones 
Pesqueras, Universidad de Montevideo, Montevideo, Uruguay, for his most help- 
ful efforts to facilitate the experiments there. I am grateful to Peter H. Klopfer 
for critically reading the manuscript. 

MATERIAL AND METHODS 

In order to avoid the bias of the hand-operated device originally inaugurated 
by Kramer (1952d), and also used by von St. Paul (1953), Hoffmann (1954; 
1960), Rawson (1954), and modified by Rawson (unpubl.) and Schmidt-Koenig 
(1958; I960), a semi-automatically operating apparatus for the directional training 
of pigeons was designed (Fig. 1). 

Twelve pecking discs, attached to micro switches, are mounted symmetrically 
at the periphery of a circular cage, one meter in diameter. Landmarks are screened 
by a removable aluminum wall. A. food cup whose cover can be opened and 
closed either manually or by a solenoid is inserted into the center of the floor. 
This portion of the apparatus rests on a base of bakelite upon which it can be 
rotated. 

The base contains 12 electrical contacts, each of which is connected to an 
electromagnetic counter mounted in a separate box. Also mounted in this box is 
a 6-volt battery, a relay to operate the solenoid and a time delay to keep the food 
cup open for several seconds at a time. The contact (i.e., the compass direction) 
which is to activate the solenoid that opens the cover of the food cup for rewarding 
a correct choice can be selected by a rotary switch. Each pecking disc is wired 
to a brush which rests on the contacts in the base. Upon proper rotation of the 
upper portion (by hand) only that particular disc, the direction of which coin- 
cides with the desired training direction, can activate the rewarding mechanism. 
In addition, each disc activates a counter corresponding to the direction to which 
it points. The entire apparatus is portable so as to facilitate long-distance 
transport. 

After 34 days of starvation, the birds, first, were taught to walk to and peck 
at a disc and to walk back and look for food in the center cup. This took a 

1 Birds from the Wilhelmshaven strain have been imported to and are heing bred at Duke 
University. Differences in the orientational anilities of the two strains are the suhject of 
another paper. 



SUN COMPASS ORIENTATION OF PIGEONS 313 

pigeon about two hours to learn. Then the bird had to learn to peck at a disc in 
one particular compass direction (the training direction). Only correct choices 
activated the rewarding mechanism. This took a pigeon several weeks to learn. 
During training the cage was irregularly rotated; during the final stages of train- 
ing, rotation was according to a random number table (the Rand Corporation, 
1955) with a range of 1-10 or 5-15 positions. For the training, the apparatus 
and the birds were transported to various open places in the immediate vicinity of 




FIGURE 1. Perspective section view of the training apparatus. Six of the twelve black 
pecking discs are shown above their sockets which house the micro switches. Netting, through 
which the birds peck at the discs, lines the circular framework around the discs ; other netting 
covers its top. In the center of the upper platform the food cup is indicated. Underneath the 
upper platform, brushes ( hatched ) are shown in contact with metal rings centered on the base 
through which the rewarding is operated. Six of the twelve contacts are depicted in the base 
(Nos. 6-11) and two of the twelve brushes subtending from the upper platform to contacts 
Nos. 6 and 11. For more details see text. 



Duke University where the apparatus was set up on top of a pickup truck. At 
other times the birds were housed in an aviary at the Duke Forest lofts. 

Five pigeons, offspring of the Wilhelmshaven strain bred at Duke University, 
were subjected to directional training on May 2, 1961 ; the training continued 
through mid-September, 1961, with 3-4 training sessions per week. The training- 
direction was to the south. The training time varied between 8:45 and 14:45 
true local time (Fig. 2). In each session, each bird was allowed to work for 
about 20 minutes. From June 15, 1961, on, each bird had to perform 5 unre- 
warded choices, from July 1 on, 10 unrewarded choices, and from August 1 on, 



314 KLAUS SCHMIDT-KOENIG 

20 unrewarded choices before the rewarded training started. The direction of the 
mean vector of these unrewarded choices from each session after July 24, 1961, 
when the training seemed to have taken shape, are given in Figure 2. 

In the critical tests in South America the birds had to perform 10-20 unre- 
warded choices in each session. This took, usually, less than 10 minutes. The 
cage was rotated irregularly between choices. A little while after the choices, the 
birds were allowed to feed from the hand-operated cup. There was no training in 
South America. 

The mean vector of all choices of each session has been calculated according 
to Gumbel, Greenwood and Durand (1953). For statistical evaluation, the 
procedure and tables of Greenwood and Durand (1955) and Durand and Green- 
wood (1958) and graph derived from these (Schmidt-Koenig, 1961. appendix), 
respectively, have been used to discriminate between random and non-random sam- 
ples. Non-random samples (/> -=0.05) were plotted in black symbols (Figs. 2-3) ; 
those random above the $% level in open symbols. If statistics could not be ap- 
plied, due to small sample size (n < 6), an open symbol with a central point is 
plotted. Unfortunately, no specific method to compute the confidence limits of 
mean vectors is yet available. 

Solar and experimental data have been calculated and plotted with reference to 
true local time (TLT), allowing for the equation of time. All solar data have 
been taken from the Nautical Almanac for 1961 and the Tables of Computed 
Altitude and Azimuth (see references). 

EXPERIMENTS AND RESULTS 
Durham, N. C. 

Figure 2 gives the performance of the birds between July 25, and September 14, 
1961, at Durham, 36 00' N; 78 56' W. Three seasonally characteristic sun 
azimuth curves are also depicted in Figure 2. All birds compensated rather well 
for the local sun's movement. The majority of choices falls into the range for 
spring and summer ; however, the scatter is too large to decide whether or not the 
birds are able to allow for the specific seasonal rate of change of azimuth of the sun. 

The translocation experiment was planned for the fall equinox of 1961. This 
would provide the least change in day /night ratio and would permit tests under the 
zenith sun. Unfortunately, political events in Brazil delayed the travel until the 
end of September. The birds were trained the last time at Durham on September 
14, 1961. From September 25 on, they were accommodated in a covered trans- 
portation crate and prevented from directly seeing the sun except during their 
individual testing periods. 

Bel cm, Brazil 

Birds and apparatus were flown to Belem during the night of September 27, 
1961. In Belem, the apparatus was set up on top of a flat-roofed building at the 
Tnstttuto Agronomico (1 27' S; 48 25' W) which had no tall landmarks around 
it. A time shift of about two hours clockwise was involved in the translocation. 
The adjustment of the birds' internal clock is known to take less than two days 
under natural conditions. 



SUN COMPASS ORIENTATION OF PIGEONS 



315 



t 



K. 



CD 




60- 



90- 




I 

Ki 



-270' 



16 18 

TRUE LOCAL TIME 

FIGURE 2. The performance of the 5 birds at Durham, N. C, between July 24 and September 
14, 1961. Each symbol represents the direction of the mean vector of all unrewarded choices 
(n = 10-20) during one examination session performed at the time of day indicated on the 
abscissa and plotted as angle to the actual sun azimuth position (left coordinate). The curves 
for the summer solstice and for the equinoxes are drawn as solid lines ; the birds' training 
fell in this period of the year. 

Birds No. 1, No. 5, and No. 6 were chosen for the equatorial tests. No bird 
reacted at all in the first testing session on September 29 at 11 :50 to 12:22 TLT. 
From September 30 on, No. 1 and No. 6 cooperated rather well when placed into 
the apparatus, exceptionally up to three times per day. If the birds were re- 
luctant to peck, this may, at least to some extent and particularly during the first 



316 



KLAUS SCHMIDT-KOENIG 



days of examination, have been due to the frequent appearance of vultures in the 
sky, which obviously frightened and immobilized the birds. No. 5, however, was 
already very unstable in its readiness to cooperate during its training in Durham. 
It turned to rather erratic choices in Belem as can be seen from Figure 3. The 
total number of successful examinations of all birds may be taken from Figure 3. 



I 



U. 
Q> 



U. 

Uj 




- 270 ( 



12 



74 



16 18 

TRUE LOCAL TIME 



FIGURE 3. The performance of pigeon No. 1, No. 5, and No. 6 at Belem, Brazil, between 
September 30 and October 10, 1961. The local sun azimuth curves are drawn for these dates. 
Symbols and way of plotting as in Figure 2. An open symbol was plotted for random choices. 



SUN COMPASS ORIENTATION OF PIGEONS 317 

The last examination was performed on October 10, 1961. Figure 3 also gives 
the local sun azimuth curves for the first and the last day of examinations. 

During the forenoon (neglecting pigeon No. 5), the bird's reference to the 
azimuth position of the sun conforms with that at Durham rather than with the 
actual local condition; however, the scatter is very large. The majority of choices 
performed during the afternoon was well off the usual range. The direction 
actually indicated by the birds shifted to the west. Hasler and Schwassmann 
(1960) seem to have encountered a similar phenomenon, although the choices of 
their fish tended to be at random rather than pointing in particular, however 
"wrong," directions. It should be indicated at this point that in the Montevideo 
tests, all birds compensated for the sun's movement much more in accordance with 
their performance at Durham, also during the afternoon. Pigeon No. 5 worked even 
more reluctantly than at home. In addition, its few choices were rather erratic. Its 
scores, therefore, are neglected in view of the more decisive performance of the 
other birds. 

The day of zenith culmination was missed, as pointed out earlier. However, 7 
tests of pigeon No. 1 and No. 6 were held when sun altitudes were more than 80. 
The respective scores of these tests, which fell within local noon time 35 minutes, 
can easily be located in Figure 3. Only in one instance (pigeon No. 6 at 12 :08 at a 
mean sun altitude of 87) were the choices (n = 9) at random with /> > 0.1. The 
other scores that are at random (pigeon No. 1 at 12:16) are due to small sample 
size (n = 2). In the remaining 5 tests at sun altitudes of 81-84 only two (pigeon 
No. 6 at 12:18 and pigeon No. 1 at 12:27, both at sun altitudes of 82) conform 
with the expectation. Thus, since the whole pattern of scores is rather widely 
scattered, the precise range of sun altitudes at which the birds are unable to derive 
compass directions from the sun has yet to be established. 

Montevideo, Uruguay 

Birds and apparatus were flown from Belem to Montevideo on October 12, and 
13, 1961. Tests were performed from October 16 through November 2, 1961, in 
an open field east of Montevideo at 34 53' S ; 56 05' W. The time shift involved in 
the translocation was less than one hour counterclockwise. There was ample time 
for the birds to become synchronized with the local day. Pigeons No. 3 and No. 7 
were principally called upon ; however, despite their extensive tests at Belem, 
No. 1 and No. 6 turned out to be still cooperative, as may be seen from Figure 4. 
Towards the end of the Montevideo time, however, the readiness of all birds to 
choose was nearly exhausted. The total number of successful tests may be taken 
from Figure 4. Also plotted in Figure 4 is the mirror image of the sun azimuth 
curve for Montevideo of October 24, 1961, the medium date of examinations and 
the azimuth curve for Durham for the same date. The true azimuth curve for 
Montevideo would fall outside the main figure and is, therefore, plotted in the 
inserted figure on a full 360 scale. 

Again, the birds clearly did not allow for the actual sun movement. The pattern 
of scores conforms more nearly to that at Durham (Fig. 2) ; however, the variance 
appears to be considerably larger. Also, there seems to be a trend to the west 
around noon and during the afternoon but far less than in the Belem tests. This 
westward shift may have been caused by the fact that the training at Durham never 



318 



KLAUS SCHMIDT-KOENIG 



90' 



60' 



A 



30< 



I 



^: o 

^ 



^ 

fc 



60' 



90 ( 




J I I I I 



MONTEVIDEO 
OCT. 2k, 1961 



90 



270 



| I I I I I/ I I I I I 

6* 9 12 15 18 
TRUE LOCAL TIME 




si 



180 



K 



210 



270' 



8 



12 



16 TRUE LOCAL TIME 



FIGURE 4. The performance of the 5 birds at Montevideo, Uruguay, between October 16 
and November 2, 1961. Also plotted in the main portion is the sun azimuth curve for Durham 
for October 24, 1961 (solid line), as the medium date of examinations and the mirror image of 
the corresponding curve for Montevideo (broken line). The actual sun curve for Montevideo 
is depicted on a 360 scale in the inserted figure. Symbols and way of plotting as in "Figures 
2 and 3. 



extended beyond 15:00, while testing at Belem and Montevideo extended until 
1 7 :00, involving westerly sun positions. A tendency to follow the sun rather than 
to maintain a fixed direction and increase the angle to the sun is frequently observed 
under conditions unfamiliar to the animal. 



SUN COMPASS ORIENTATION OF PIGEONS 319 

DISCUSSION AND CONCLUSION 

From the scores of the pigeons it can be safely concluded that they did not allow 
for the equatorial or trans-equatorial sun movement. The birds seem to have 
largely referred to the sun as if it were the sun at Durham. The scores in South 
America, though mostly significantly non-random, appear to be more widely spread 
than those at the training place. It remains unanswered whether this may be 
interpreted as a specific reaction resulting from the disagreement between orienta- 
tional features expected by the bird and those actually observed, or as a nonspecific 
result of handling the bird in unusual and strange conditions, however irrelevant 
to orientation, or simply as a fading of the training. Nevertheless, the short period 
of exposure to the local sun of about 10 minutes was clearly insufficient to allow the 
bird to take full acount of the discrepancies between local solar parameters and 
those at home. Nonetheless, migratory birds or the navigating pigeons should be 
able to take account of these within that time period if the sun is the basis for 
navigation. Previous experiments on the sun compass in northern latitudes failed to 
demonstrate solar altitude as a source of information for directional choices in 
starlings and pigeons (Hoffmann, 1954; Schmidt-Koenig, 1958). A number of 
specifically designed homing experiments did not support solar altitude as providing 
navigational information to pigeons (Kramer, 1953, 1955, 1957; Rawson and 
Rawson, 1955; Hoffmann, 1958; Schmidt-Koenig, 1958, 1961). 

The homing pigeon, a nonmigratory species, was chosen for this initial series, 
because of its technical advantages. That it is not an entirely irrelevant subject 
for studies of sun compass orientation in equatorial regions may best be illustrated 
by a non-specific account : it is illegal to keep homing pigeons in Brazil because they 
are extensively used for smuggling diamonds and narcotics. This speaks for 
their reliability in homing, at the very least. 

A new series of experiments is under way in which pigeons, raised and trained 
at Durham, N. C.. will be exposed to the local conditions in South America in an 
aviary several months in advance of tests. This procedure may reveal whether 
individuals are able to adjust to grossly different solar conditions. The present 
series may serve as a control experiment. 

In the translocation experiments with bees (Kalmus. 1956; Lindauer, 1959, see 
also 1960) the examinations concerned offspring of displaced individuals rather 
than the translocated individuals themselves. Kalmus (1956) claims that even the 
offspring did not adjust to the new solar conditions. Lindauer (1959) clearly 
demonstrated in his experiments that the offspring allowed for the equatorial sun 
movement. This apparent controversy has not been settled yet. Another relevant 
experiment was concerned with displaced fish (Hasler and Schwassmann, 1960). 
The authors consider their data as preliminary and exploratory and only drew 
tentative conclusions. It is unfortunate that the majority of the scores of their 
fish so far published are in fact not very convincing, due to a sample size just too 
small for statistical evaluation and to too large variance. This is, in turn, clue to 
specific difficulties arising from work with fish. 

SUMMARY 

1 . Five homing pigeons were directionally trained at Durham, N. C. (36 00' N ; 
78 56' W). An automatically recording and rewarding cage was used. The 



320 KLAUS SCHMIDT-KOENIG 

directional response of the birds was tested upon displacement to Belem, Brazil 
(1 27' S; 48 25' W) and subsequently to Montevideo, Uruguay (34 53' S; 
56 05' W). Tbe birds were prevented from directly seeing the sun except for 
the actual test periods of about 10 minutes on each occasion. 

2. The birds clearly did not allow for the respective local sun movements in 
South America but referred to the sun as if it were the sun at Durham. The 
choices in South America were more widely scattered than those at home. 

LITERATURE CITED 

BIRUKOW, G., 1956. Lichtkompassorientierung beim Wasserlaufer ] 7 clia currcns F. (Hetero- 

ptera) am Tage und ztir Nachtzeit. Zcitschr. Ticrpsychol., 13: 463-484. 
BRAEMER, W., 1959. Versuche zu der im Richtungsgehen der Fische enthaltenen Zeitschatzung. 

Verh. Dt. Zool. Gcs. 1959. Miinster Westf. 276-288. 
BRAEMER, W., 1960. A critical review of the sun-azimuth hypothesis. Cold Spring Harbor 

Symposia, 25: 413-427. 
DURAND, D., AND J. A. GREENWOOD, 1958. Modification of the Rayleigh test for uniformity in 

analysis of two-dimensional orientation data. /. Geology, 66: 229-238. 
FISCHER, K., 1961. Untersuchungen zur Sonnenkompassorientierung und Laufaktivitat von 

Smaragdeidechsen (Laccrta viridis Laur.). Zcitschr. Ticrpsychol., 18: 450-470. 
VON FRISCH, K., 1950. Die Sonne als Kompass im Leben der Bienen. Expericntia, 6: 210-221. 
GOULD, E., 1957. Orientation in box turtles, Tcrapene c. carolinensis Linnaeus. Biol. Bull., 112: 

336-348. 
GREENWOOD, J. A., AND D. DURAND, 1955. The distribution of length and components of the 

sum of n random vectors. Ann. Math. Stat., 26: 233-246. 
GUMBEL, E. J., J. A. GREENWOOD AND D. DURAND, 1953. The circular normal distribution : 

theory and tables. /. Amer. Stat. Assoc., 48: 131-152. 
HASLER, A. D., R. M. HORRALL, W. J. WISBY AND W. BRAEMER, 1959. Sun-orientation and 

homing in fish. Limn, and Occanog., 3: 353-361. 
HASLER, A. D., AND H. O. SCHWASSMANN, 1960. Sun orientation of fish at different latitudes. 

Cold Spring Harbor Symposia, 25: 429-441. 
HOFFMANN, K., 1954. Versuche zu der im Richtungsfinden der Vogel enthaltenen Zeitschatzung. 

Zeitschr. Ticrpsychol., 11: 453-475. 

HOFFMANN, K., 1958. Repetition of an experiment on bird orientation. Nature, 181 : 1435-1437. 
HOFFMANN, K., 1960. Experimental manipulation of the orientational clock in birds. Cold 

Spring Harbor Symposia, 25: 379-387. 
KALMUS, H., 1956. Sun navigation of Apis mellifica L. in the southern hemisphere. /. E.vp. 

Biol., 33: 554-565. 
KRAMER, G., 1952a. Die Sonnenorientierung der Vogel. Verli. Dt. Zool. Ges. 1952. Freiburg 

Brsg. 72-84. 

KRAMER, G., 1952b. Experiments on bird orientation. Ibis, 94: 265-285. 
KRAMER, G., 1953. Wird die Sonnenhohe bei der Heimfindeorientierung verwertet? /. Ornithol., 

94:201-219. 
KRAMER, G., 1955. Ein weiterer Versuch, die Orientierung von Brieftauben durch jahreszeitliche 

Anderung der Sonnenhohe zu beeinflussen. Gleichzeitig eine Kritik der Theorie des 

Versuchs. /. Ornithol., 96: 173-185. 

KRAMER, G., 1957. Experiments on bird orientation and their interpretation. Ibis, 99: 196-227. 
LACK, D., 1959. Migration across the sea. Ibis, 101 : 374-399. 

LACK, D., 1960. Migration across the North Sea studied by radar. Ibis, 102: 26-57. 
LINDAUER, M., 1959. Angeborene und erlernte Komponenten in der Sonnenorientierung der 

Bienen. Zcitschr. vergl. Physiol., 42 : 43-62. 
LINDAUER, M., 1960. Time-compensated sun orientation in bees. Cold Spring Harbor Symposia, 

25: 371-377. 

NAUTICAL ALMANAC FOR THE YEAR 1961. United States Naval Observatory, Washington, 1959. 
PARDI, L., AND F. PAPI, 1952. Die Sonne als Kompass bei Talitrus saltator Montagu (Amphi- 

poda Crustacea). Natunviss., 39: 262-263. 






SUN COMPASS ORIENTATION OF PIGEONS 321 

PERDECK, A. C, 1958. Two types of orientation in migrating starlings, Sturnus vulgaris L. and 

chaffinches, Fringilla coelebs L., as revealed by displacement experiments. Ardea, 46: 

1-37. 
RAND CORPORATION, 1955. A million random digits with 100,000 normal deviates. The Free 

Press, Glencoe, 111. 
RAWSON, K. S., 1954. Sun compass orientation and endogenous activity rhythm of the starling 

(Sturnus vulgaris L.). Zeitschr. Tierpsychol., 11: 446-452. 
RAWSON, K. S., AND A. M. RAWSON, 1955. The orientation of homing pigeons in relation to 

change in sun declination. /. Ornitlwl., 96: 168-172. 

ROWAN, W., 1946. Experiments in bird migration. Trans. Roy. Soc. Canada, 40: 123-135. 
RUPPELL, W., 1944. Versuche iiber Heimfinden ziehender Nebelkrahen (Corvus cor one comix) 

wahrend des Wegzuges. /. OrnithoL, 92: 106-132. 
RUPPELL, W., AND E. SCHUZ, 1948. Ergebnisse der Verfrachtung von Nebelkrahen (Corvus 

corone comix) wahrend des Wegzuges. Vogehvarte, 15: 30-36. 
VON ST. PAUL, U., 1953. Nachweis der Sonnenorientierung bei nachtlich ziehenden Kleinvogeln. 

Behaviour, 6: 1-7. 
ScHMiDT-KoENic, K., 1958. Expcrimentelle Einflussnahme auf die 24-Stunden-Periodik bei 

Brieftauben und deren Auswirkungen unter besonderer Beriicksichtigung des Heim- 

findevermogens. Zeitschr. Tierpsychol., 15: 301-331. 
ScHMiDT-KoENic, K., 1960. Internal clocks and homing. Cold Spring Harbor Symposia, 25: 

389-393. 
ScHMiDT-KoENic, K., 1961. Die Sonne als Kompass im Heim-Orientierunssystem der Brief- 

taube. Zeitschr. Tierpsychol, 18: 221-244. 
TABLES OF COMPUTED ALTITUDE AND AZIMUTH I AND IV. US Navy Hydrographic Office, 

Washington 1958. 



SODIUM, POTASSIUM AND CHLORIDE IN SELECTED HYDROIDS 

H. BURR STEINBACH * 

Department of Zoology, University of Chicago, Chicago 37, Illinois 

The survival of small, relatively simple animals, in aquatic habitats ranging 
from fresh waters to marine, poses problems of ion- and water-regulation that still 
need to be defined in quantitative terms. It has been shown that flatworms, acoelo- 
mate "organ-level" forms of wide distribution, appear to have a characteristic 
ionic content which persists in the face of ionic variations in the environment 
(Steinbach, 1962a). K content of organisms tends to remain in characteristic 
amount and high with respect to the environment ; Na and Cl are maintained above 
environmental levels in fresh-water forms, and below environmental levels in 
marine forms with intermediate conditions in various degrees of dilution of sea 
water (Steinbach, 1962b). 

While the flatworms have achieved an "organ-level" of organization, the 
coelenterates may be characterized as "metazoa of the tissue grade of construction" 
(Hyman, 1940). Of the coelenterates, the members of the class Hydrozoa range 
from the fresh-water hydras, through such brackish-water forms as Cordylophora 
to the many colonial hydroids in marine habitats. Typically marine hydrozoa 
(Topping and Fuller, 1942) appear to have had little success at colonizing the 
variable brackish waters of estuaries. On the other hand, Cordylophora is wide- 
spread in brackish- to fresh-water ponds and Hydra occurs commonly throughout 
the fresh-water habitat. 

The growth, general behavior and distribution of Cordylophora has been re- 
ported extensively (Roch, 1924; Kinne, 1958; Fulton, 1961). Much special in- 
formation about the hydroids is to be found in the symposium report edited by 
Lenhoff and Loomis (1961). Titbitlaria has been widely used for studies of re- 
generation. However, the only basic information about water and salt composi- 
tion and the mobility of these constituents in Hydrozoa is contained in a paper by 
Lilly (1955). Lilly demonstrates, in Hydra, an influx of Na, K and Br and a 
permeability to water, and gives figures for Na and K content based on an assumed 
isotopic equilibrium. 

The present studies were undertaken to determine directly the Na and K con- 
tents and obtain influx values for Hydra in dilutions of Ringer's solution with pond 
water, and of Cordylophora and Tubularia in dilutions of sea water. 

It is shown that the actual analytical content of H. littoralis - (the species used 

1 Work aided by grants from the National Science Foundation (#12449) and the Wallace 
G. and Clara A. Abbott Memorial fund of the University of Chicago. 

2 Both Hydra littoralis and Chlorohydra riridissiina were originally obtained from local 
supply houses and then cultured in this laboratory. No expert taxonomic identification was 
made but the specimens certainly fit the descriptions in a general way (cf. Hyman, 1931). 
Chlorohydra viridissima is probably very similar if not identical with the European Hydra viridis 
(Hyman, 1929). In general the habitat of Pelmatohydra oligactis is very similar to that of 
Hydra littoralis. It may be assumed that Tubularia crocea was used although there does seem 
to be some doubt about the identity of specimens from Woods Hole and from Cape Cod Canal. 

322 



Na, K AND Cl IN HYDROIDS 

here) is at the least double that inferred for Pelmatohydra by Lilly, and indicates 
very slowly exchangeable components of both Na and K. 

Insofar as the present studies on Hydra littoralis are concerned, it is assumed, 
with the exceptions noted, that the experimental findings of Lilly on Pelmatohydra 
oligactis apply. On this basis, Hydra littoralis is assumed to ( 1 ) maintain a con- 
stant volume in osmolar concentrations of the external medium ranging from pond 
water to about 40 millimolar sucrose, (2) possess a marked permeability to water 
as evidenced by shrinkage in higher concentrations of sucrose, and (3) show an 
appreciable rate of exchange of Na and K contents with the environmental 
constituents. 

MATERIALS AND METHODS 

4 

Hydra littoralis - was cultured according to the general procedures outlined 
by Loomis (cf. Lenhoff and Loomis, 1961) in a medium containing Versene, Ca 
and bicarbonate in Chicago tap water. They were fed Arternia nauplii five times a 
week, except the subcultures set aside for experimental use. In general. Hydra 
were not fed for at least one day prior to analysis. A few observations were made 
on Chloroh\'dra znridissima cultured in Chicago tap water. 

Numerous Hydra were shaken loose from their attachments in the culture dish 
and pipetted with culture medium into previously tared conical, 12-ml. centrifuge 
tubes. The tubes were then centrifuged 5 minutes at ca. 1600 X gravity and the 
supernatants decanted. The weight of the "Hydra pellet" could then be determined. 
Carboxyl C 14 inulin was added in some instances to give some estimate of "non- 
Hydra" space. This "space" presumably includes a portion representing non- 
drainage of medium from the tube, medium around the animals and possibly me- 
dium in the coelenterons of the animals. The specimens of Hydra were not 
crushed by the gentle centrifuging, Chlorohydra individuals, especially, acting like 
rather stiff bristles and the "pellet" showing a correspondingly greater "inulin 
space" than the pellet of H. littoralis. The results are reported as concentrations 
in the centrifuged pellets, inulin values being indicated to show the magnitude of 
the corrections that could be made on the assumption that inulin spaces represented 
space outside the animals. 

Cordylophora lacitstris, obtained originally from Cape Cod, was cultivated ac- 
cording to the general methods discussed by Fulton (1961), using dilutions of natu- 
ral sea water from Woods Hole. General growth characteristics of Cordylophora 
in various dilutions of sea water have been reported by others (cf. Kinne, 1968) ; 
the behavior of the cultures used in these experiments was not markedly different 
from that described. Growth was very slow in concentrations below 10% sea 
water and above 50</ under the conditions used. Stems with hydranths and with 
occasional bits of stolons were harvested from the appropriate cultures and im- 
mersed in the desired medium, usually identical with that in which they were 
grown. The specimens could then be placed briefly on hard filter paper, allowed 
to drain a few seconds and transferred to slips of Parafilm for rapi.d weighing on 
torsion balance. This treatment did not appear to disturb the hydranths unduly. 
Specimens so treated appeared normal five minutes after return to their normal 
medium. The results reported here for ion contents and ion fluxes refer to the 
whole stem-hydranth system without attempting to separate the contributions of 
perisarc, coenosarc, or coelenteron. 



324 



H. BURR STEINBACH 



Tubularia 3 was obtained from Woods Hole Harbor and from the north end of 
the Cape Cod Canal. There is no assurance that all were of the same species the 
classification appears to be somewhat in doubt. However, no differences were 
noted between animals from the two sources and hence no precise identification 
was attempted. 

In general, Tubularia was treated and prepared for analysis according to the 
methods outlined for Cordylophora. Tubularia hydranths are very large, plenti- 
fully endowed with gonophores during the season of these investigations and in- 
clined to drop off the stems after a day or so at sea-water-table temperatures 
(19-21 C. during the summer of 1962). Sections of stems, however, seemed 
to remain in good shape, starting to regenerate new hydranths after loss of the 
original ones. Therefore, the results reported herewith relate only to stems, rather 
than to whole stem-hydranth units. 

Analysis of ionic contents and measurement of ion fluxes involved the routine 
methods reported in other studies (cf. Steinbach, 1962a). 

TABLE I 

Ionic composition o/ Chlorohydra viridissima and Hydra littoralis. Concentrations in millimoles/kg. 

Corrected concentrations were calculated assuming imdin space is not part of the animal. 

The inulin space in parentheses was from another experiment on the same batch 

of animals. Medium in all cases <1.0 mM in salt. 





Pellet 




Corrected 






Inulin Space 






Na 


Cl 


K 


,0 


Na 


Cl 


K 


Chlorohydra 


9.3 




11.0 


47 


17 




21 




5.1 


1.5 


10.3 


66 


15 


4 


30 


Hydra 


9.5 


3.4 


24.0 


37 


15 


5 


38 




18.0 


5.7 


35.0 


(30) 


26 


8 


50 



Na and K were assayed by flame photometer on dilute hot acid extracts of the 
animals. Cl was determined by a Cotlove-type chloridemeter, and Na 24 and K 42 , 
used to study ion fluxes, were assayed with a thin window or Nal well counter, 
corrections for radioactive decay being made where necessary. 

RESULTS 
A. Hydra 

Table I gives the analytical figures for two determinations on Chlorohydra 
viridissima and two on H. littoralis made during the same season. While Chloro- 
hydra pellets have lower ionic concentrations than H. littoralis, "correction" for 
inulin spaces brings the values closer together. In view of the complexity of 
Chlorohydra, with its symbiotic algae, the rest of the results reported will concern 
H. littoralis. 

Although no tests for acclimation to increased salinity over long times were 
attempted, H. littoralis can withstand concentrations up to 20% of amphibian 

3 See footnote 2. 



Na, K AND Cl IN HYDROIDS 



325 



TABLE II 

Ionic contents of pellets of H. littoralis following exposure of the animals for 30 minutes or more to 

diluted frog Ringer's solution. Concentrations in millimoles per liter, or per kilogram with 

the number of analyses averaged in parentheses. Average inulin space in 

comparable experiments = 31%. 



Solution 


Pellet 


Na 


Cl 


K 


Na 


Cl 


K 


0.2 





<0.1 


14(3) 


6(3) 


32 (3) 


6.0 


7.2 


<0.1 


14(5) 


10(5) 


35 (5) 


9.0 


9.6 


2.0 


18(5) 


12(5) 


26(5) 


20.0 


22.0 


2.5 


24(4) 


17(4) 


38(4) 



Ringer's solution. Loomis (1959) reports normal growth in 5% sea water. 
Table II gives the analytical results on Hydra held half an hour or more in solu- 
tions of the composition indicated. There is a remarkable constancy of K con- 
centrations, regardless of either total ionic strength or K content of the bathing 
medium. Na and Cl contents both increase with increasing external concentration 



TABLE III 

Entry of Na into Hydra littoralis. Influx expressed as counts /minute /kg. pellet. Na concentra- 

trations: mM/kg. pellet or medium. Per cent exchange calculated as specific activity of 

pellet -f- specific activity of medium X 100. Medium: dilute frog Ringer's. 



Immersion time 


Na" 


Na (total) 


Per cent exchange 


1.00 hour 


3.9 


35 


37 


3.15 hours 


3.7 


23 


53 


18.50 hours 


4.1 


19 


73 


44.00 hours 


4.3 


18 


80 


Medium 


6.0 


20 


100 



TABLE IV 

influx. Pellet of H. littoralis. Solutions of composition as indicated in millimols/ liter. 

Figures as in Table III. 





K 


K 


Per cent exchange 


Solution 


297 


0.9 


100 


Na = 1, K = 0.9 








6 hours 


1630 


26 


19 


24 hours 


4230 


31 


42 


Na = 9, K = 0.9 








6 hours 


1730 


37 


14 


24 hours 


4450 


43 


31 


24 hours 


5050 


45 


34 



326 



H. BURR STEIN BACH 




20 40 60 80 100120140160180200220240 
Cone. NaCI Medium 

FIGURE 1. Concentrations (mAf/kg.) of Na, Cl and K in Cordylophora, determined as 
described in the text, plotted against ionic strength of the medium expressed as concentration of 
NaCI (mM/L). Points represent single determinations except those marked "H" = values for 
Hydra from Table II. Straight lines have a slope of 0.75. 



Na, K AND Cl IN HYDROIDS 



327 



hut iii a manner indicating only partial distribution in the pellet. Presumably tbc 
volume of the animals did not change appreciably upon transfer to solutions in 
the osmotic range used (see Lilly, 1955). 

Na 24 and K 4 " both penetrate H. littoralis, but slowly (Tables III and IV). Re- 
sults to date are not precise enough to give reliable rate figures but the half-times 
for exchange with Na and K of the whole pellets are in the range of 15 to 30 
hours, the longer time being indicated for K. The tables also give evidence of a 
rapid early entrance of Na- 4 , which may be due to entry of Na- 4 into the coelenteron 
or other extracellular areas. In one experiment on penetration of Na 24 from a very 
dilute pond-water medium, there was some evidence of a nearly completely un- 
exchangeable fraction of Na. 

B. Cordylophora 

Cordylophora specimens allowed to grow in various dilutions of sea water were 
analyzed. The results are summarized in Figure 1. Na and Cl content vary 
directly with the environment, the general trend indicating a possible equilibration 
with 50-60% of the volume of the animals. K, on the other hand, remains re- 
markably constant in concentration, regardless of the external ionic strength. 
Whether a constant concentration of K also means a constant amount per cell it 
is not possible to say, since relative weights of the animals in different media were 
not determined. As the sea water dilutions were made with distilled water, the K 
concentration of the environment also varied from a high of about 5 millimolar to 



K 




2345 
Time - Hours 

FIGURE 2. Semi-log plot of exchange of Na 24 and K 42 , medium to Cordylophora against time. 
Open circles, NaCl concentration of medium = 140 mM, half-filled circles, NaCl concentration 
= 80 mM. Dots = K influx from medium 4 mM K and 140 mM Na. Ordinates = %X 0.1. 



328 



H. BURR STEINBACH 



TABLE V 

Nti, Cl and K concentrations (mM/kg.) of Tubularia stems immersed in solutions indicated for 4 or 

more hours. Dilutions of sea water with distilled water. All concentrations expressed 

as averages of 5 determinations with standard errors of mean indicated. 



Medium 


Na 


Cl 


K 


Sea water 
75% sea water 
50% sea water 


283 db 21 
206 12 
105 14 


345 23 
253 11 
165 7 


71 1 

54 2 
47 1 



less than one millimolar. Therefore, the internal K concentration reflects neither 
a constant inside/outside ratio nor a variation representing osmotic adjustment. 

Influx of Na 24 or K 4 - was measured by removing a standard sample of Cordylo- 
phora stems at intervals after addition of the appropriate isotope to the medium, 
blotting gently on hard filter paper and counting directly under a thin-window 
counter. The animals survived this handling well, most of the hydranths appearing 
intact after a half dozen or more measurements during the three- to five-hour 
period of the experiment. At the end of the period indicated, the selected samples, 
and other samples not subjected to the repeated handling, were extracted, and the 
extracts analyzed for the ionic constituents and radioactivity as usual. On the 
basis of the specific activities so determined, the points plotted in Figure 2 were 
determined. 



100 



80 



o 

UJ 

o 



I 

o 

X 

UJ 



60 



40 



20 



-O 



468 

TIME-HOURS 



10 



FIGURE 3. Influx of Cl 30 into Tubularia stems. Circles : influx from propionate lobster 
Ringer's (Cl = 77 mM) . Dots : influx from normal sea water (Cl = 520 mM) . 



Na, K AND Cl IN HYDROIDS 



329 



K 4 - exchanges slowly from a medium containing 5 millimolar K and 140 milli- 
molar Na. An estimated minimum half-time for exchange would he of the order of 
eight hours. During the three-hour period indicated, influx appears to follow a 
simple kinetic curve, disregarding a very small initial rapid uptake. The time period 
was not sufficient to indicate whether or not there was an apparently unexchangeable 
fraction of K. Na 24 exchanges more rapidly (half-time ca. 4 hours for the slow 
phase) and shows a very rapid initial influx. There is some indication that there is 
a portion, about 25%, of the total Na that is very slowly exchangeable. Influx of Na 
is at nearly the same rate from 80 millimolar and 140 millimolar external Na 
concentrations. 

C. Tubularia 

Because of the large size and relative delicacy of the hydranths of Tubularia, the 
results reported will relate only to analyses of stem segments. The data in general. 



100 
80 



O O- -O., __ 



- - -o 




8 12 16 

TIME -HOURS 



20 



24 



FIGURE 4. Influx of Na 24 into Tubularia stems. Circles : from choline chloride artificial sea 
water (MBL formula), Na = 25 mM. Dots : from normal sea water, Na = 440 mM. 

however, apply to hydranth material, the major differences being those that can be 
accounted for by the smaller inulin or sulphate space for hydranths (17%) as com- 
pared to stems (28%). The perisarc of the stems may be regarded as essentially 
inert with respect to passage of materials. A few experiments in which empty peri- 
sarc tubes were loaded with K 42 sea water showed nearly complete equilibration with 
an outside medium in five minutes. Insofar as possible with the available material, 
straight clean stem segments were used. However, some of the colonies, especially 
later in the season, had extensive branching patterns. 



330 



H. BURR STEINBACH 



100 

80 - 

O 

60 - 



40 



20 



C\J 

1- 

O 
UJ 



S 10 

x 

2 8 

ID 



8 12 16 

TIME-HOURS 



20 



24 



FIGURE 5. K 42 influx into Tubularia from normal sea water (dots) and sea 
water made 60 mM with added KC1 (circles). 



Na, K AND Cl IN HYDROIDS 



331 



Table V gives average- values for contents ot I'lihiilnrid stems maintained in 
normal sea water, and in 75 '/< and 50 r / dilutions of sea water with distilled water. 
A few measurements indicated only slight swelling of the tissues; thus, the ionic 
changes probably represent movement of the respective elements rather than dilution 
effects due to water intake. 

Na and Cl concentrations of Tubularia stems are quite variable, as indicated by 
the standard errors of the means. K concentration is markedly more constant in the 



360 





340 Y 




320 

mf 




300 f 




280 


- 


- 


260 


- 


3 


240 


- 


CO 

ZD 


220 


- 


1 


200 


- 


O 

or 
o 

o 


180 
160 




CJ 

O 
o 


140 
120 


Na 




100 


9 ' ci 




80 
60 


- />/ 






40 
20 




i i i i i i i 


1234567 



TIME - HOURS 

FIGURE 6. Concentration of Na or Cl (m;l//kg.) in Tubularia stems transferred at zero 
time to Cl-poor medium (propionate lobster Ringer's) or Na-poor medium ( choline chloride sea 
water). Horizontal labelled lines indicate concentrations of respective ions in medium. Time: 
hours immersion in experimental fluids. 



332 



H. BURR STEINBACH 



Concentrations of AV/, Cl and K and % exchange of AV/- 1 in 'l'ul)iilaria stems held at sea water table 

temperature (ca. 20 C.) for the times indicated. Concentrations and exchange values 

expressed as in Table III. Alcohol-treated stems removed at 24 hours, 

immersed in #5% ethanol 30 seconds and then returned 2 hours 

to Na' M sea water. 



Time 


Na 
cone. 


', Xa 
exchange 


Cl 
cone. 


K 
cone. 


13 hours 


258 


76 


332 


85 


24 hours 


194 


87 


276 


62 


32 hours 


213 


93 


291 


81 


48 hours 


193 


103 


280 


81 


Alcohol-treated 


368 


100 


444 


<10 



different batches of stems analyzed. In a few cases, Na 2 S 35 O 4 or C 14 inulin was 
added to the equilibration medium. As with the other hydroids, assay for S 3S or C 14 
after an hour or so equilibration indicated a "space" of about 30%. 

Na, K and Cl are all exchangeable, at least in part, as indicated by influx of the 
appropriate isotopes into Tubularia stems (Figs. 3, 4, 5). 

Cl 26 penetrates rapidly at first from normal sea water, the initial phase involving 
more of the total Cl than can be accounted for on the basis of 30 % extracellular space 
(Fig. 3 ). Testing influx from lobster Ringer's fluid made with sodium propionate 4 
instead of the normal NaCl (Cl == 77 mM rather than 520 mM) has the major 
effect of decreasing the initial rapid phase, the slower delayed phase of penetration 
(t. -= 5-6 hours) being about normal. There is, of course, a net loss of Cl (Fig. 6) 
during the equilibration period, the major change again occurring rapidly. Cl of 
Tubularia appears to exist mainly in freely diffusible form but involving at least 
two "compartments" which do not appear to be equated to sulphate and non-sulphate 
space. 

TABLE VII 

Effect of excess KCl on exchange of ions; medium to Tubularia stems. Exchange calculated as in 

Table III with appropriate isotope measured. K-sea water = 5% M/l. KCl. Time: 

6-7 hours. Single determinations of paired runs for each isotope. 



Exchange 



Na 24 



Normal sea water 
K-sea water 


75 
65 


80 
76 


35 
74 



Na 2 * penetrates from normal sea water (Fig. 4) with a rapid initial phase, 
mostly accounted for by penetration into sulphate space, followed by a very slow 
phase (t} = > 15 hours). Long-term experiments showed complete equilibration 
at the end of 48 hours (Table VI) ; stems immersed 30 seconds in absolute ethanol 
and then returned to the radlioactive medium took less than two hours (shortest 
time measured) for complete equilibration. 

4 I am indebted to Dr. Harry Grundfest for a supply of this medium. 



Na. K AND Cl IN HYDROIDS 



333 



Penetration of Na 24 from an artificial sea water made with choline chloride in- 
stead of NaCl (Na == 25 mM instead of 440 mM') shows a loss of the first rapid 
phase of penetration and a slower phase (tj - 40 hours). Immersion of Tubu- 
laria stems in choline sea water leads to an initial rapid net loss of Na ( Fig. 6 ) , 
the tissue concentration at the end of 6 hours still being notably above that of the 
medium. 

K 42 penetration from normal sea water appears to be relatively uncomplicated. 
The experiments do indicate a fast, followed by a slow influx into the tissues. 
Penetration of K 4 - from sea water with excess KC1 (K == 60 mM instead of 
10 mM) indicates the drastic speeding of the slow phase. This effect of excess K 
of the medium is specific for influx of K. Excess K of the medium has little 
effect on the influx of Na or Cl (Table VII ) . 

Raising the K concentration of the medium increases that of the tissue slightly to 
a new level which is then maintained (Table VIII). Assuming a 30% extra- 

TABLE VIII 

Effect of immersion of Tubularia stems in normal 95% sea water and 95% sea water plus 5% by 

volume M/l. KCl. Figures in parentheses indicate number of separate 

determinations in averages. 



Immersion time 


Na 


Cl 


K 


Normal Sea Water 
<1 hour 
4-7 hours 


276 (2) 
258 (4) 


331 (2) 
320 (4) 


72 (2) 
71 (4) 


60 mM KCl Sea Water 
< 1 hour 
4-7 hours 


255 (2) 
245 (4) 


330 (2) 

328 (4) 


90 (2) 
92 (4) 



cellular space, cell concentration is raised only from 97 mM to 105, the Kj/K ratio 
changing from approximately 10 to 1.7. 

DISCUSSION 

A striking fact emerging from the present studies is the relative constancy of K 
concentration in the hydroids investigated, ranging from the fresh-water Hydra 
littoralis in pond water and diluted frog Ringer's, through the brackish-water 
Cordylophora lacustris in various dilutions of sea water, to the strictly marine 
Tubularia. The results, to be sure, represent analyses of whole organisms with 
contributions of extracellular materials (perisarc, mesogloea and cavities) and un- 
drained fluid of pellets or stems. However, C 14 inulin studies indicate a free dif- 
fusion space of the order of 30% for all types of animals for short immersions, and 
the probability is that correction for extracellular space, while increasing the 
concentration values, would not destroy the uniformity. 

The only other report known to me on K concentration of the hydroid phase 
of hydrozoans is that of Lilly (1955) who inferred concentrations in Pelmatohydra 
from influx curves for Na 24 and K 42 . While it is possible that the low values she 
reported represent species differences, it seems probable that her assumption of 



334 H. BURR STEINBACH 

complete equilibration with external K 42 is responsible. Whether the very slow 
rate of equilibration of intracellular K with externally applied isotope indicates any 
real "binding" cannot be proven from the data. The flux data are reminiscent 
of the situation found for K influx into whole frog muscle (cf. Harris, 1957). 

Further studies are needed to determine water fluxes. However, assuming that 
the high permeability to water deduced from osmotic experiments on H \dra (Lilly, 
1955) is a condition common to the hydroids, then their osmotic and ionic problems 
would appear to be as follows : 

1. In the range, pond water to 40 millimolar sucrose (ca. 5% sea water), 
Hydra volume remains constant. Growth rates are similar in pond water and 5 c /c 
sea water (Loomis, 1959). The environmental osmotic range for constant body 
volume may be much broader for Cordylophora. 

2. There is a remarkable relative constancy of K concentration of the animals in 
all external salt concentrations tested. Hydra living in pond water (NaCl equiva- 
lent <0. 1 mM) has over half the K concentration of Tnbnlaria from sea water 
( XaCl equivalent >500 mJ/ ). This indicates a K uptake mechanism and a K-reg- 
ulating mechanism not primarily controlled by internal/external ratios nor by total 
ionic strength of the medium. 

3. With Hydra, and Cordylophora in very dilute sea water, Na and Cl are 
slightlv in excess of the environmental concentrations. At higher environmental 
salt concentrations, Na and Cl appear to penetrate freely into about 60% of the 
volume of the animal. The findings with H\dra indicate an Na uptake mechanism 
in dilute solutions. There may be an Na extrusion mechanism in concentrated salt 
media. If, however, there is an ion extrusion mechanism, it is a rather odd one, 
behaving as though it could deplete about half the cell volume of the animal of Na 
and Cl regardless of the external concentration. The situation has points of 
similarity to that described for Tetrahymena (Dunham and Child, 1961 ). 

The behavior of K seems clear and unequivocal. There is a K concentration 
(and probably an amount of K per unit cell type) that is fixed and is maintained 
regardless of external ionic strength or without relevance to a particular Kj/K,, 
ratio. 

It has been suggested that, widespread throughout the animal kingdom, there is 
an optimal internal K concentration on the order of 150 millimolar (Steinbach, 
1962b). Fresh-water invertebrates have long been known to be more dilute forms 
with respect to tissue K concentration. While careful correction of the whole- 
animal K concentrations found in the hydroids for "extra-cellular" space would give 
higher values for cellular K than for whole-animal K, it is doubtful that they would 
more than approach 100 millimolar. It will be of interest to see whether there is 
a characteristic "minimal level" of K concentration necessary for the continued 
existence of animal cells living even in the most dilute media. 

There is as yet no conclusive evidence of any extensive "binding" (= immobiliza- 
tion) of ions of the alkali metals in protoplasm. Analvsis of flux rates in a variety of 
forms has indicated some heterogeneity of both cellular Xa and cellular K but on 
the basis of the data currently available, it must be assumed that the cellular Na and 
K of any living system are diffusible and constitute a major portion of the osmotic 
concentration of the cells. It thus seems likely that the internal osmotic pressure 



Na, K AND Cl IX HYDKOIDS 335 

of Hydra cells is even greater than the apparent isotonic point of 40 millimolar 
sucrose indicated by the shrinkage of the animals in higher concentrations (Lilly, 
1955). On the basis of the chemical analyses reported here, the true osmolar 
concentration would be nearly double the figure cited. 

Without in any way intending to suggest that cellular mechanisms for ionic 
uptake and regulation of internal contents are understood, the water balance of 
the fresh-water forms presents the most challenging picture. With H\dra, for 
example, water balance might conceivably be achieved by a very impermeable outer 
layer with a filtration-resorption system in the endoderm-lined gastrovascular cavity. 
This seems unlikely, both in view of Lilly's studies on tentacles of Pclinato- 
liydra and in view of the fact that the Hydra can survive yawning periods 
of up to half an hour with, presumably, free diffusion access of the gastro- 
vascular cavity to the very dilute environment. This hydroid system deserves 
careful study as a possible example of a tissue system showing water transport 
without excretory systems on the organ level nor constant cellular organelles such 
as contractile vacuoles. On the other hand, hydroid tissues are richly endowed 
w^ith vacuoles of various sizes ( Yf. papers in "The Biology of Hydra," Lenhoff and 
Loomis ) which may prove to be concerned with water regulation. 

It is suggestive that fresh-water hydroids as well as flatworms, coping with 
strongly hypotonic environments, are endowed with extensive mucus-secreting sys- 
tems. A relatively thick mucous layer, perhaps merely by providing an "unstirred" 
external layer, is probably of critical importance in osmotic protection, although it 
seems unlikely that such a layer would act as a regulatory device. 

It was hoped that the studies reported here would offer precise suggestions about 
ecological factors controlling the distribution of the hydroids, the occurrence of the 
three types used being quite different. The hydras in general appear limited to 
fresh water, though they will survive and feed in at least 5% sea water (Loomis, 
1959). Titbularia is a typical marine form of a type not penetrating usually into the 
brackish waters of estuaries (Topping and Fuller, 1942). 

Cordylophora appears to occur only in brackish water, not. or rarely, in the 
very dilute fresh water nor in full-strength ocean salt water (Roch, 1924). Cordy- 
lophora is the most versatile of the forms reported here since it can thrive and re- 
generate in a variety of salinities. Its continued growth has also been shown to be 
independent of any specific ion ratios, although there is a specific requirement for 
the presence of Na, K and Cl (Roch, 1924). Reconstitution of extruded tissue 
masses requires Ca or Mg and K (Beadle and Booth, 1938). 

Titbnlaria and related forms will live and regenerate in sea water of f { ocean 
strength or less (rf. Keil, 1932) even though they are not usually found in such 
dilute habitats. A point of interest is that many marine invertebrates are reported 
to regenerate better in dilute than in full-strength sea water (cf. Keil, 1932). 

With these different habitat restrictions for the hydroids, it would be good if 
different methods of handling ionic and osmotic problems could be demonstrated. 
Unfortunately, the results reported here offer no suggestions. Ionic and osmotic 
conditions appear to be reflected in a general pattern of distribution common to the 
hydroids rather than showing special variations related to habitat. A possible ex- 
ception would be the presence of an Xa and Cl uptake mechanism from very dilute 
environments. Ability of hydroids to survive in high salt media probably reflects 



336 H. BURR STEINBACH 

an ability of the metabolic systems to function in high Na environments, rather than 
the possession of mechanisms to hold the tissue Na concentration to low levels. 

The tendency to uniform tissue K concentration may be assumed to be general 
for all cell types of the hydroids. Thus, Tulnilaria hydranths, richly endowed with 
nematocysts, musculo-epithelial cells and nervous elements, have much the same 
electrolyte pattern as Tnbiilaria stems which consist, exclusive of perisarc, of 
mostly ectodermal and endodermal cells without further specialization (Hyman, 
1940). 

LITERATURE CITED 

BEADLE, L. C., AND F. A. BOOTH, 1938. The reorganization of tissue masses of Cordylophora 

Iticnsfris and the effect of oral cone grafts with supplementary observations on Obcha 

(/clatinosa. J. E.vp. Biol.. 15: 303-326. 
DUNHAM, P. B., AND F. M. CHILD, 1961. Ion regulation in Tetrahymena. Biol. Bull., 121: 

129-140. 
FULTON, C. H., 1961. The Development of Cordylophora. In: The Biology of Hydra. ( H. M. 

Lenhoff and W. F. Loomis, eds.) University of Miami Press, Coral Gables, Florida. 
HARRIS, E. J., 1957. Permeation and diffusion of K ions in frog muscle. /. Gen. Ph\sioL, 41: 

169-195. 
HYMAN, L. H., 1929. Taxonomic studies on the hydras of North America. I. General remarks 

and description of Hydra aincricana. new species. Trans. Atner. Mic. Soc., 48: 242-255. 
HYMAN, L. H., 1931. Taxonomic studies on the hydras of North America. IV. Description of 

three new species with a key to the known species. Trans. Amcr. Mic. Soc., 50: 302-315. 
HYMAN, L. H., 1940. The Invertebrates. I. Protozoa through Ctcnophora. McGraw-Hill Book 

Co., Inc., New York. 
KEIL, E. M., 1932. The effect of salts upon the regeneration of Pcnnarla tiarclla. J . E.vp. ZooL, 

63:447-455. 

KINNE, O., 1958. Adaptation to salinity variations: some facts and problems. In: Physiologi- 
cal Adaptation. (C. L. Prosser, eel.) Washington, D. C., Amer. Physiol. Soc. Pp. 

92-106. 
LENHOFF, H. M., AND W. F. LOOMIS, 1961. The Biology of Hydra. University of Miami Press, 

Coral Gables, Florida. 

LILLY, S. J., 1955. Osmoregulation and ionic regulation in Hydra. J. Exp. Biol., 32: 423-439. 
LOOMIS, W. F., 1959. Control of sexual differentiation in H\dra by pCO 2 . Ann. N. Y. Acad. 

Sci.. 77: 73-86. 
ROCH, F., 1924. Experimented Untersuchungen an Cordylophora caspia (Pallas) [= lacustris 

Allman] iiber die Abhangigkeit ihrer geographischen Verbreitung und ihrer Wuchs- 

formen von den physikalischchemischen Bedingungen des ungebenden Mediums. 

Zeitschr. Morph. Okoloi/ic Ticrc. 2: 350-426. 

STEINBACH, H. B., 1962a. Ionic and water balance of planarians. Biol. Bull., 122: 310-319. 
STEINBACH, H. B., 1962b. The prevalence of K. Perspectives in Biol. Mcd., 5: 338-355. 
TOPPING, F. L., AND J. L. FULLER, 1942. The accommodation of some marine invertebrates to 

reduced osmotic pressures. Biol. Bull., 82: 372-384. 



THE REGENERATION OF WHOLE POLYPS FROM ECTODERMAL 

FRAGMENTS OF SCYPHISTOMA LARVAE OF 

AURELIA AURITA 1 

SONIA N. STEINBERG 2 
Biology Department, Brandeis University, Waltham 54, Mass. 

Gilchrist (1937) reported that isolated ectodermal fragments of scyphistoma 
larvae of Aurelia anrita reconstituted an entire organism. The report did not in- 
clude a histological study and gave no indication of the cell type involved in the 
formation of the new endodermal layer. Gilchrist conjectured that interstitial cells 
were responsible for the formation of the new layer. In view of the uncertainty of 
the role of interstitial cells in hydroid regeneration and the fact that some features 
of reconstitution may occur in their absence (Brien ct al., 1953; Normandin, 
1960), Gilchrist's experiments with scyphistomae of Aurelia were repeated and an 
attempt was made to trace the origin of the endoderm by means of a histological 
study of a staged series of reconstituting ectodermal isolates. 

The observations here presented indicate that Aurelia scyphistomae have no 
cells which conform to the characteristics of typical hydroid interstitial cells. A 
count of mitoses reveals that somatic ectodermal cells are the most actively dividing 
cells and that they give rise to a population of cells (amoebocytic in appearance) 
which in turn become the new endodermal cells. 

MATERIALS AND METHODS 

Scyphistoma larvae of the marine scyphozoan, Aurelia anrita, were obtained 
through the courtesy of Dr. Sears Crowell, and were grown in the laboratory at 
room temperature. They were kept in small fingerbowls or plastic Petri dishes 
containing filtered sea water obtained from Woods Hole, Mass. The larvae were 
fed each morning for 30 to 60 minutes on freshly-hatched brine shrimp (ArtemiaJ. 
The Artcinia were hatched in sea water, rinsed several times, and introduced into 
the cultures of Aurelia. After the animals had fed, the culture water was replaced 
with freshly filtered sea water. Water was changed again each evening. 

Animals were removed from the stock culture, placed in Syracuse dishes con- 
taining filtered sea water, and starved for 24 hours prior to operation. 

All operations were performed with the aid of a dissecting microscope at a 
magnification of either 15 X or 20 X. Peduncle, hypostome and tentacles were re- 
moved with sharpened surgical needles. The remaining cylinder of tissue was slit 

1 Based on a thesis submitted in partial fulfillment of requirements for a Master's degree in 
Biology at Brandeis University. Supported by grant G-6334 from the National Institutes of 
Health to Dr. E. Zwilling of Brandeis University. 

2 Present address : Laboratory of Neuroanatomical Sciences, National Institute of Neuro- 
logical Diseases and Blindness, National Institutes of Health, Bethesda 14, Maryland. 

337 



338 



SONIA N. STEINBERG 




FIGURE 1. Cross-section through the mid-body region of Aitrclla, showing ectoderm (on 
left), endoderm, mesoglea and an amoeboid cell. (X 840. ) 

FIGURE 2. Cross-section of an ectodermal fragment six hours after its separation from the 
endoderm, showing several amoeboid cells in the lumen. ( X 520. ) 

FIGURE 3. Cross-section of a two-day-old regenerating ectodermal fragment. A distinct 
endoderm is present, and an amoeboid cell can be seen between the two layers. (X 520.) 



REGENERATION OF AURELIA ECTODERM 339 

lengthwise, flattened, and oriented so that the endodermal layer was facing upward. 
The tissue was then cut into six strips, each about 0.5 mm. wide, and the length of 
the original cylinder (about 1 mm.). The rectangular piece of tissue was held 
with a #5 watchmaker's forceps while a cut was made to the level of the mesoglea 
with a surgical needle. The endoderm was separated from the ectoderm by 
teasing and cutting through the mesoglea. 

In Amelia, the ectoderm is unpigmented, while the endoderm contains an orange 
pigment. This pigment is a useful diagnostic marker which helps to insure the 
elimination of all contaminating endoderm. The pigment marker, plus the thick 
mesoglea which separates the two layers, enables one to be certain of isolation of 
pure ectoderm. The endoderm-free isolates were transferred in a pipette to plastic 
Petri dishes containing filtered sea water. Fragments were sacrificed immediately 
after the operation (time 0), at 2, 4, 6, 8, 10, 12, 18, 20, 22, 24, 30, and 36 hours, 
and at 2, 3, 4, 5, and 6 days. 

The specimens were fixed in one of the following fluids : Bouin's, Gilson's pre- 
pared with sea water, Carnoy's, Kleinenberg's, and Zenker's acetic. Bouin's and 
Gilson's were most useful, and were used in the majority of the work. 

Animals were fixed for one-half to one hour, dehydrated in a graded ethanol 
series, cleared in cedarwood oil, and embedded in either 60-62 C. mp paraffin with 
10% beeswax, or in 56-58 C. mp paraffin. Sections were cut at 7 microns. 

Several stains were employed: toluidine blue (0.25% aqueous solution at pH 5), 
iron haematoxylin, Feulgen's, Azure B, Mallory's triple aceto-orcein. Only the 
first three successfully stained the sections of Amelia, and they were used in most 
of the experiments. 

RESULTS 
Preliminary data 

Thirty-two isolated ectodermal fragments were cultured in filtered sea water. 
All of these fragments formed complete hydranths. Twelve started to capture 
freshly hatched Artemia on their sixth day, the other 20 on the seventh. Ingestion 
of the shrimp began one day later. In addition, 41 specimens, all having a longest 
dimension of at least 0.18 mm., were raised until they had formed at least one 
tentacle. Tentacles began to appear six or seven days after isolation of the ecto- 
derm. Eleven additional fragments, all less than 0.18 mm. long, that did not show 
any signs of developing tentacles by day seven, were scored as negative and were 
discarded. 

Normal histology of Aurelia 

The intact scyphistoma of Aurelia aurita is a two-layered, sac-like animal. In 
the outer, single-celled layer of ectodermal cells there are various epithelial deriva- 
tives : epitheliomuscular cells, cnidoblasts, nerve cells, columnar epithelium. The 
inner endodermal layer is also one cell thick. Some endodermal cells are digestive, 
some glandular, and others muscular. The endoderm is invaginated into the 

FIGURE 4. Longitudinal section through the mouth of a six-day-old regenerating specimen. 
The endodermal layer is complete, hasopliilic glandular cells surround the mouth, and a tentacle 
has formed. (X235.) 



340 SONIA N. STEINBERG 

lumen to form four longitudinal ridges, which enclose the gastric pouches. The 
two cell layers are separated by an acellular mesogleal layer (Fig. 1). 

The base of the animal is further differentiated into a holdfast region in which 
the endodermal cells are enlarged and elongated. The apical region contains a 
mouth which is surrounded by a dense population of small, basophilic glandular 
cells in the endodermal layer. The mouth is ringed by a circle of tentacles, each of 
which bears several batteries of nematocysts. 

Histology of regenerating ectodermal fragments 

The development of an entire polyp from an ectodermal fragment from the mid- 
body region (see Fig. 1) was followed at intervals from to 7 days. 

(a) hours: Fourteen specimens were fixed immediately after the ectoderm 
was separated from the endoderm. They consist of fragments of ectoderm, and 
some adhering mesoglea. The ectoderm appears as a single folded layer of columnar 
cells. Endoderm is not present in any of the sectioned specimens. An occasional 
large amoeboid cell is in the mesoglea of some specimens, but in no cases at a fre- 
quency of greater than 1 per 250 cells. Each explant contains a few thousand cells 
which are in a compact, though fragmented, mass. 

(b) 2 hours: The three explants examined have lost their fragmented appear- 
ance, and have developed into a hollow mass of cells. The shape of the mass is ir- 
regular, but all of its contours are rounded. The inner margin of the ectoderm is 
lined with a thin layer of mesoglea. 

(c) 4 hours: There is a slight increase in the number of amoeboid cells in the 
mesoglea of the three specimens examined. Some ectodermal cells appear to be 
necrotic. An occasional mitotic figure is visible in the outer margin of the columnar 
ectodermal layer. 

(d) 6 hours: Fifteen specimens were fixed. Mitotic figures are more numer- 
ous in the ectoderm than at 4 hours, and there is an increase in the number of amoe- 
boid cells. The latter cells are confined to the lumen, and are attached to the 
mesoglea. They exhibit long thin processes, are vacuolated, and have a large ec- 
centric nucleus. No mitotic figures have been seen in them. The ectodermal cells 
are still all columnar, and arranged in an orderly fashion, similar to that observed in 
the normal animal (Fig. 2). 

In some areas there is a necrotic mass containing many refractile inclusions and 
semi-lunar, Feulgen-positive bodies. The refractile bodies appear to be breakdown 
products of cnidoblasts. Mitotic activity is not found in the area surrounding the 
necrotic masses, and amoeboid cells are rare in this region. 

(e) 8 hours (3 specimens) : There are areas in which the lumen is obliterated 
by necrotic masses. In non-necrotic portions of the animal there are invaginations 
of ectoderm ; the population of amoeboid cells is most dense in these regions of 
invagination. 

(f) 10 hours (3 specimens) : The lumen is large, and the areas of necrosis are 
less numerous. The number of amoeboid cells is about the same as at six hours. 

(g) 12 hours (14 specimens) : Mitotic activity is still confined to the ectoderm, 
and is occurring at a surprisingly high rate of about 1/250 cells. There is a marked 
increase in the amount of mesoglea. 



REGENERATION OF AURELIA ECTODERM 341 

(h) 18 hours (11 specimens) : The mitotic rate is more than double that seen 
at 12 hours. Mitotic activity is confined almost exclusively to the ectoderm; in all 
of the sections studied only one amoeboid cell was seen in the process of division. 

(i) 20 hours (3 specimens) : No apparent change is visible. 

(j) 22 hours (3 specimens) : Amoeboid cells are numerous, and are found 
scattered throughout the lumen. Several are packed densely near the point of in- 
vagination of the ectoderm. The ectoderm is now extremely regular, with the long 
axis of each cell pointing to the center of the enclosed area and the nuclei all basal. 
Cnidoblasts have appeared. 

(k) 24 hours (18 specimens) : Some areas of the lumen are densely populated 
with amoeboid cells. The mitotic rate has not changed noticeably since 18 hours, 
and mitotic activity is still confined to the ectoderm. 

(1) 30 hours (2 specimens') : The amoeboid cells are beginning to line up end 
to end, or to arrange themselves into small groups. They are confined to the lumen, 
and are visible on both sides of the mesogleal layer, with which they are in contact. 
Some cells have extended long filamentous processes toward neighboring cells, and 
are in contact with these neighbors via the resulting bridges. There is little 
mitotic activity. Necrotic areas are no longer visible. 

(in) 36 hours (2 specimens) : The amoeboid cells are lined up in tandem, and 
form a distinct second layer, in which some typical endodermal cells may be seen 
for the first time. 

(n) 2 days (11 specimens) : There are two distinct layers of cells; the outer is 
ectodermal, the inner is endodermal. Mitotic activity is confined to the ectoderm 
and is becoming more frequent (Fig. 3). 

(o) 3 days (13 specimens) : Endodermal mitoses are becoming apparent, but 
are less frequent than ectodermal mitoses. An endodermal layer is distinct and, 
with the exception of one specimen, amoeboid cells are very rare. The exceptional 
specimen has one area with abundant amoeboid cells and no endodermal layer. 
Except for the one area this specimen has a distinct endodermal layer lacking 
amoeboid cells. 

In the other specimens at this time the endoderm is a distinct, coherent layer, 
and, in usual fashion, is separated from the ectoderm by the mesoglea. A mouth is 
beginning to form at one end of the mass. 

(p) 4 days (12 specimens) : There is little change from three days. Amoeboid 
cells are rare in all specimens. 

(q) 5 days (6 specimens) : The animals are forming apical outpocketings (con- 
sisting of both ectoderm and endoderm) in the region of the hypostome. There are 
imaginations of the endoderm which resemble gastric pouches. 

(r) 6 days (3 specimens) : The mouth has formed, and the area of the 
hypostome is surrounded by small, basophilic glandular cells. At the opposite end 
of the animal, the endodermal cells have enlarged into typical vacuolated stalk 
cells. Tentacles, complete with batteries of nematocysts, are present, although 
they vary in number and size from specimen to specimen. Mitotic activity is found 
in both ectoderm and endoderm ( Fig. 4 ) . 

(s) 7-10 days: Some time during this period the regenerant begins to feed. 
Several specimens, including at least one from each set of experiments, have been 
followed for a few weeks, until they bud off a new scyphistoma. 



342 SONIA N. STEINBERG 

DISCUSSION 

Gilchrist (1937) assumed, with no microscopic evidence to support his con- 
tention, that the newly formed endodermal cells were formed from interstitial cells 
in his ectodermal isolates of Aurelia. The present study has failed to reveal typical 
interstitial cells in either an intact scyphistoma, or during any stage of reconstitu- 
tion. During regeneration, large amoeboid cells, similar to ones seen only infre- 
quently in the intact animal, begin to appear. These cells do not have the typical 
basophilic character of a standard interstitial cell ; instead, they are undistinguished 
cells, with no definite shape or characteristic cytoplasmic constitution. 

These amoeboid cells could have come from either pre-existing amoeboid cells 
or from somatic ectoderm cells. The evidence indicates that the latter is the case. 
This evidence is based on fourteen ectodermal isolates which were selected at 
random during the experiments and fixed immediately after preparation. Nine of 
these contained no amoeboid cells ; the other five had such cells but in no case was 
there more than one for every 250 ectoderm cells. Since all isolates of pure ecto- 
derm which were larger than 0.18 mm. in length regenerated completely, we can 
assume (if the fourteen isolates are representative) that initial presence of amoeboid 
cells is not required for formation of an endodermal layer. Further, there was a 
consistent increase in the relative number of amoeboid cells with time after isola- 
tion, and such cells were found in all specimens fixed four hours or longer after 
preparation. Since this increase occurred in the absence of significant mitotic ac- 
tivity in the amoeboid cell population and was accompanied by a high mitotic 
rate of the ectoderm, we are led to the conclusion that the amoeboid cells, which 
eventually form the endodermal cells, are derived from the somatic ectoderm cells. 

The source of cells in regeneration has not been clearly established. Attempts 
to link the new cells to interstitial cells or neoblasts often involve x-irradiation. 
Such treatment inhibits migration of interstitial cells, and regeneration ceases 
(Brien and Reniers-Decoen, 1955; Burnett, 1961; Puckett, 1936). One can con- 
clude that interstitial cells are necessary for regeneration only if one assumes that 
x-irradiation does not affect somatic cells. Most studies of regeneration involve 
cases where all somatic cell types are represented. These are known to divide, and 
can be the source of the new cells. The question at issue is whether a somatic cell 
of one type is capable, regardless of its pathway, of forming a cell of another type. 
It has been proposed that, once a cell has received and used the information which 
determines its adult type, it loses its capacity to become another type of cell. The 
situation can best be assessed where one tissue type is eliminated. This has been 
done in the present study. Aurelia possesses no reserve cells, and in the present 
study it has been shown to be capable of reconstituting an endodermal layer when 
this layer is absent. The evidence is strong that somatic ectoderm can give rise 
to somatic endoderm, by first losing its characteristics and becoming an "indifferent" 
amoeboid cell. 

SUMMARY 

1. Isolated ectodermal fragments of Aurelia aurita scyphistomae regenerate 
into complete hydranths. Aurelia possesses no interstitial cells, but large amoeboid 
cells appear during reconstitution. 



REGENERATION OF AURELIA ECTODERM 343 

2. The endoderm forms by differentiation of the amoeboid cells, which have 
probably arisen by dedifferentiation of the ectoderm. 

LITERATURE CITED 

BRIEN, P., AND J. P. VAN DEN EECKHOUDT, 1953. Bourgeonnement et regeneration chez les 

Hydres irradiees par les rayons X. C. R. Acad. Sci., 237: 756-758. 
BRIEN, P., AND M. RENIERS-DECOEN, 1955. La signification des cellules interstitielles des hydres 

d'eau douce et le probleme de la reserve embryonnaire. Bull. Biol. France Belgique, 

89: 258-325. 

BURNETT, A. I., 1961. The growth process in Hydra. J. Exp. Zoo/., 146: 21-83. 
GILCHRIST, F. G., 1937. Budding and locomotion in the scyphistomas of Aurelia. Biol. Bull 

72: 99-124. 

NORMANDIN, D. K., 1960. Regeneration of Hydra from the endoderm. Science, 132: 678. 
PUCKETT, W. O., 1936. The effects of x-radiation on the regeneration of the hydroid, Pennaria 

tiarella, Biol. Bull, 70: 392-399. 




CUTANEOUS AND PULMONARY GAS EXCHANGE IN THE 
SPOTTED SALAMANDER, AMBYSTOMA MACULATUM 1 

WALTER G. WHITFORD AND VICTOR H. HUTCHISON 

Department of Zoology, University of Rhode Island, Kingston, Rhode Island 

The first quantitative study of pulmonary and cutaneous respiration in amphib- 
ians was conducted on Rana esculenta and R. fitsca by Krogh (1904). He inserted 
a cannula, connected to an air pump, into the trachea and analyzed separately the 
air forced through the lungs by the pump and the air surrounding the frog. Krogh 
found that carbon dioxide was released chiefly through the skin, while oxygen was 
taken up predominantly by the lungs. He also found that oxygen uptake through 
the skin remained relatively constant throughout the year, while oxygen uptake 
by the lungs was greatest during the spring and dropped below cutaneous uptake 
during the fall and winter. His curves for release of carbon dioxide through the 
skin and lungs followed the pulmonary oxygen uptake curve throughout the year. 
Dolk and Postma (1927), using similar techniques with Rana tcmporaria, sub- 
stantiated Krogh's results. 

Lapicque and Petetin (1910) demonstrated that cutaneous respiration in the 
lungless salamander, Euproctus montanus, may be more important than lung and/ 
or buccopharyngeal respiration. They found that E. montanus dies quickly when 
submerged in Vaseline with its head free, but can live without buccopharyngeal 
respiration. However, their study did not solve the problem of the relative im- 
portance of cutaneous and buccopharyngeal or pulmonary respiration in salamanders. 

Since the relative roles of pulmonary and cutaneous respiration in salamanders 
had not been studied quantitatively, we undertook the present study to determine 
the role of each in the respiration of Ambystoma maculatum. 

METHODS AND MATERIALS 

The animals used in this study were collected in late March, 1962, in the vicinity 
of Kingston, R. I. Groups of animals were acclimated in constant temperature 
chambers in total darkness. The minimum standards used for acclimation were 
as follows : 

5 C. one week at 15 C. ; two weeks at 10 C. ; 10 days at 5 C. 

10 C. one week at 15 C. ; two weeks at 10 C. 

15 C. one week at 15 C. 

25 C. one week at 15 C.; one week at 25 C. 

30 C. one week at 15 C. ; one week at 25 C.; 5 days at 30 C. 

A mask of 0.5-inch Tygon flexible plastic tubing was sutured to the head of the 
animal at least 24 hours prior to use. The mask was constructed in a manner that 

1 Supported by National Science Foundation Grant 13953. 

344 



GAS EXCHANGE IN A SALAMANDER 



345 



would not interfere with nnrni.'d huccal movements (Fig. 1). Pulmonary and 
cutaneous respiration were determined in a closed system respirometer. 

The respirometer consisted of four chambers of equal volume constructed of 
0.25-inch acrylic plastic ( Fig. 2 ) . The two rear chambers served as thermo- 
barometers, while pulmonary and cutaneous respiration were measured separately 
and simultaneously in the two front chambers, which were connected by a 0.5-inch 
hole. A cover of 0.25-inch plastic was screwed down over the four chambers and 
sealed with petroleum jelly. A series of plastic connectors with stopcocks com- 
municated through the cover into the respiration chambers. Manometers with 
colored kerosene indicators were fitted into the stopcock connectors. Syringes 
filled with 100% oxygen were fitted to the stopcock connectors of the pulmonary 





FIGURE 1. Method of attaching plastic mask to salamanders. 

and cutaneous chambers. All connections were checked for leaks prior to each set 
of experiments. 

The respirometer was constructed within a larger plastic chamber which served 
as a water jacket. A cooling coil, heating coil, temperature regulator, and stirrer 
were placed in the water jacket for temperature control. The temperature-regu- 
lating system kept the water temperature constant within 0.1 C. 

The masked animal was tied firmly to a piece of hardware cloth and the mask 
fitted through the hole between the chambers. The end of the mask was sealed 
into the hole in the chamber wall by the application of petroleum jelly. Beakers 
containing 10 ml. of barium hydroxide were placed in each chamber to absorb 
carbon dioxide. The beakers of barium hydroxide contained plastic-coated mag- 
netic bars which were moved at regular intervals by a magnet outside the 
respirometer chambers. This stirred the barium hydroxide solution and insured 
an effective absorption of the carbon dioxide by breaking the barium carbonate 



346 



WALTER G. WHITFORD AND VICTOR H. HUTCHISON 



film which funned on the surface and which would have resulted in reduced ab- 
sorption of carbon dioxide. Oxygen injected into the chambers by the syringes 
compensated for oxygen consumed by the animal. Oxygen consumption was read 
directly from the calibrated syringes. 

At the end of a set of experiments, the beakers of barium hydroxide were re- 
moved from the chambers and titrated with standardized 1 N sulfuric acid to 
determine the quantity of carbon dioxide produced. The beakers of barium hy- 
droxide in the thermobarometers served as controls, since each beaker of barium hy- 
droxide absorbed carbon dioxide at the same rate both prior to the experiment and 
during the time required for titration. To determine the actual amounts of carbon 
dioxide released by the animal, the amount of carbon dioxide absorbed in the 



2 



TR 



TB 




WJ-- 



B S ^ 



10 cm 



FIGURE 2. Apparatus used to measure simultaneously pulmonary and cutaneous respiration 
in amphibians. AM, animal mask ; B, barium hydroxide solution in beaker ; C, cooling coil ; 
H, heating coil ; M, manometer ; Mg, magnetic stirring bar ; O 2 , oxygen syringe ; TB, thermo- 
barometer chamber ; TR, temperature regulator ; S, stirrer ; WJ, water jacket. 

thermobarometers was subtracted from the amounts of carbon dioxide absorbed in 
the pulmonary and cutaneous chambers. 

The few animals that struggled against their bonds during the first hour of the 
experiment produced high oxygen consumption values. The values obtained for 
these hours were not included in the calculations of mean oxygen consumption, but 
the carbon dioxide produced had to be included in the determination of respiratory 
quotients, since the barium hydroxide could not be removed for titration after 
each hour. 

Measurements of total oxygen consumption were made as controls. Differences 
in oxygen consumption between masked and unmasked animals were not statisti- 
cally significant. Oxygen consumption measurements for masked and unmasked 
animals tied to hardware cloth indicated that restraint of the animals resulted in a 



GAS EXCHANGE IN A SALAMANDER 347 

slight increase in oxygen consumption. Total oxygen consumption for experi- 
mental animals was about 5% higher than unrestrained controls at all tempera- 
tures except 5 C., where movement was negligible. Respiratory quotient values 
for control animals were not significantly different from those of experimentals. 

Tidal volumes were measured by connecting the animal's mask to a graduated 
manometer. The volume of air required to move the manometer column a dis- 
tance equal to that moved by the breathing of the animal was taken as the tidal 
volume. 

Recordings of breathing movements were obtained for masked and unmasked 
animals acclimated to 10 C., 15 C., and 25 C. by passing a loop of thread under 



60 



60 



JZ 

O) 
U1 

4> 



^40 



20 




- -_.. __ A 



10 20 30 

Temperature C 

FIGURE 3. Mean cutaneous and pulmonary gas exchange at different temperatures. 

the buccal floor of the animal and connecting it to a pressure transducer. Move- 
ment of the buccal floor caused movement of the transducer wire, which was con- 
nected to a Physiograph. 

RESULTS 

Pulmonary and cutaneous gas exchange 

Pulmonary oxygen consumption increased almost linearly from 1.36 //l./gm./hr. 
at 5 C. to 86.59 ^l./gm./hr. at 30 C. Cutaneous oxygen increased from a mean 
of 22.74 ^l./gm./hr. at 5 C. to 73.00 ^l./gm./hr. at 15 C., then dropped to 
46.62 /xl./gm./hr. at 30 C. (Figs. 3 and 4). The ratio of pulmonary to cutaneous 



348 



WALTER G. WHITFORD AND VICTOR H. HUTCHISON 



30C (52-10) 


I | I I i 1 1 i i 1 ' 1 

C 




P 


30C (52-10) 


i I HH^H ) i 


25C(57-10) 


C 


25C(57-10) 


p 




c 


15C (55-15) 


( 1 HBIBHI 1 


15C (55-15) 


P 


10C (50-10) 


C 


10C (50-10) 


P 


5C(36- 6) 


C 


5C(36-6) 


p 
1 1 1 1 1 1 1 1 1 1 1 1 . 1 



20 40 60 80 100 

ju liters /g / hr 



12O 



140 



FIGURE 4. Cutaneous and pulmonary oxygen consumption at different temperatures. The 
first figure in parentheses indicates total number of hours of measurement; the second figure 
denotes number of individuals in sample. The horizontal line indicates the range, the thin 
vertical line, the mean ; one black and one white rectangle combined on each side of the mean, 
one standard deviation ; one black rectangle on each side of the mean, two standard errors. If the 
standard errors of two sets of data do not overlap, the difference between the means may be 
considered statistically significant (Hubbs and Hubbs, 1953). 

oxygen consumption increased with temperature (Fig. 5 ). No measureable amount 
of carbon dioxide was released through the lungs and buccopharyngeal mucosa at 
5 C. A sharp increase in both pulmonary and cutaneous carbon dioxide occurred 
between 10 and 15 C. At temperatures above 15 C. pulmonary carbon dioxide 
remained almost constant (Figs. 3 and 6). Cutaneous carbon dioxide increased 
gradually with increasing temperature. The ratio of pulmonary to cutaneous 
carbon dioxide release was approximately 0.2 at all temperatures except at 5 C.. 
where there was no measureable release of carbon dioxide from the lungs and 
buccopharyngeal surfaces. 



GAS EXCHANGE IN A SALAMANDER 



349 



Respiratory quotients 

The range and mean of respiratory quotients, (R(J), were: 5 C. 0.73-0.81, 
x = 0.76; 10 C, 0.71-0.85, x = 0.77; 15 C, 0.72-0.89, x==0.78; 25 C, 
0.70-0.87, x==0.76; 30 C., 0.71-0.88, x == 0.77. Animals in acclimation were 
fed a diet of mealworms at regular intervals; those that refused to eat had RQ 
values between 0.70 and 0.72. In animals kept at 30 C., the mean RQ changed 
from 0.82 one to three days after feeding, to 0.72 five days after feeding. For 
animals acclimated to 10, 15, and 25 C., RQ's between 0.84 and 0.80, observed 



T 



T 



3.5 



3.0 



O 

01 

8 

c 

03 
*J 

<J 



2.5 



2.0 



1.5 



OJ 

o: 



1.0 



0.5 




10 20 

Temperature C 



30 



FIGURE 5. The ratio of pulmonary to cutaneous oxygen consumption at different temperatures. 

Method of presentation is the same as in Figure 4. 



350 



WALTER G. WHITFORD AND VICTOR H. HUTCHISON 



30C (52-10) 
30C (52-10) 
25C (57-10) 
25C (57-10) 
15C (55-15) 

1 5C(55-15) 

10C (50-10) 

10C (50-10) 

5C (36- e ) 

5C (36 - 6 ) 



C 



p 

none 







20 



40 60 

CO 2 jJliters/ g / hr 



80 



100 



FIGURE 6. Cutaneous and pulmonary carbon dioxide release at different temperatures. 
Method of presentation is the same as in Figure 4. 

from one to four days after feeding, decreased to 0.72 eight days after feeding. At 
5 C., mean RQ's increased from 0.74 three days after feeding, to 0.78 five days 
after feeding. Meal worms could be detected in the stomachs of animals acclimated 
to 5 C., up to five days after feeding, indicating incomplete digestion and 
assimilation. 

Tidal "volumes and breathing rates 

Two distinct breathing movements, including two separate tidal volumes, occur 
in salamanders with lungs. The buccopharyngeal movement consists of an en- 
largement of the buccopharyngeal cavity by a lowering of the hyobranchial appara- 
tus, resulting in the inspiration of air through the nares. Exhalation occurs when 
the buccal floor rises again. A pronounced depression of the buccal floor occurs 
at intervals. During the latter part of this depression, the nares completely close, 
the buccal floor raises and forces the air into the lungs (Whipple, 1906). 

Temperature had a direct effect on tidal volumes (Fig. 7). The mean buc- 



GAS EXCHANGE IN A SALAMANDER 



351 



copharyngeal tidal volume increased from 0.008 cc. at 5 C. to 0.065 cc. at 25 C. 
and 30 C. The mean lung tidal volume increased from 0.09 cc. at 5 C. to 0.42 
cc. at 30 C. 

The number of deep inspirations, i.e., those movements in which air was forced 
into the lungs, remained relatively constant (between six and nine per minute at 



0.4 



0.3 



u 
o 



0) 

E 



0.2 



fT 



0.1 







-O Mean Lung Tidal Volume 







Mean Buccopharyngeal 
Tidal Volume 












10 



20 
Temperature C 




FIGURE 7. Mean lung and buccopharyngeal tidal volumes at different temperatures. 
Vertical lines denote the range of observed variation. 



352 WALTER G. WHITFORD AND VICTOR H. HUTCHISON 

all temperatures). However, the rate of buccopharyngeal oscillations was de- 
pendent upon temperature. The average rate of buccopharyngeal movements was 
13 at 10 C, 80 at 15 C, and 91 at 25 C. Masking had no visible effect on 
breathing rates. 

The mean volume of air inspired by buccopharyngeal movements at 10 C., 
was 15.6 cc. per minute; by lung inspirations, 68.4 cc. per minute; but at 25 C. 
the mean volume of air inspired by buccopharyngeal movements was 382.2 cc. per 
minute and by lung inspirations, 194.4 cc. per minute. Thus, the volume of air 
moved by buccopharyngeal oscillations at 25 C. is approximately 25 times that 
moved at 10 C., while that moved by the lungs is only about three times as great. 

The per cent efficiency of the lung-buccopharyngeal mucosa combination was 
calculated by dividing the oxygen consumption per hour by the total amount of 
oxygen inspired, the total volume of inspired air being calculated from breathing 
rates and tidal volumes. The following are the calculated efficiencies: 10 C., 
1.43% ; 15 C, 0.93% ; 25 C, 0.86%. 

DISCUSSION 

Krogh (1904) and Dolk and Postma (1927) probably introduced experimental 
error into their measurements of lung respiration by failure to acclimate their 
experimental animals adequately to constant temperature. In addition, they 
cannulated the trachea and forced air in and out of the lungs with a mechanical 
pump, a procedure that probably does not duplicate the two distinct breathing 
movements in frogs (Cole and Allison, 1929; Scholten, 1942). The mask used 
in our experiments did not affect either the breathing movements or the breathing 
rates of the experimental animals. 

With increasing environmental temperatures, the lungs and buccopharyngeal 
mucosa play an increasing role in oxygen uptake in A. maculatnin. Since oxygen 
uptake through the skin is passive, it is dependent upon the proximity of the 
capillaries to the surface of the skin, the rate of blood flow through the capillaries, 
and the affinity of the hemoglobin for oxygen. Uptake of oxygen through the 
lungs and buccopharyngeal mucosa is not only dependent on these same factors 
but also on the depth and rate of breathing movements. Therefore, the increase 
in oxygen uptake through the lungs and buccopharyngeal mucosa that occurs with 
rising temperature can be directly correlated with increases in tidal volumes and 
in breathing rates. At the same time, the rate of lung inspirations remains rela- 
tively constant, changing from six at 10 C. to nine at 25 C., while the rate of 
buccopharyngeal oscillations increases greatly, from 13 per minute at 10 C. to 91 
per minute at 25 C. 

Matthes (1927), Vos (1936) and Elkan (1955) concluded that the bucco- 
pharyngeal oscillations of amphibians were olfactory in function, while Noble 
(1925) had assumed that these movements were primarily for respiration. Czopek 
(1962) found that the capillaries of the mouth cavity in A. opacum accounted 
for only 6.4 r /<< of the respiratory capillaries and concluded that (p. 586), "the 
pulsation of the buccal floor ... is probably connected with olfactory functions 
rather than respiration." He pointed out, however, that (p. 586) "conclusions 
derived exclusively from morphological findings must he accepted with prudence 



GAS EXCHANGE IN A SALAMANDER 353 

unless they are supported by physiological investigations." Our data indicate that 
buccopharyngeal movements are of appreciable value in respiration, especially at 
higher temperatures. Between 10 C. and 25 C, the volume of air moved through 
the lungs increases about three-fold ; the volume of air moved through the bucco- 
pharyngeal cavity increases 25-fold. If the air moved by the buccopharyngeal 
oscillations were excluded from respiration, the efficiency of the lungs would have to 
double to account for the increased oxygen consumption at 25 C. 

In nature during the warmer months of the year, A. maculatuin would probably 
have food in its stomach three to five days after feeding. At normal environ- 
mental temperatures (10 C. to 25 C.), five to six days are necessary for com- 
plete digestion and assimilation of food because during this period RQ values 
remain above fasting levels. 

Cutaneous oxygen uptake increased linearly between 5 C. and 15 C., but 
dropped to lower values at 25 C. and 30 C. This decrease in oxygen uptake 
through the skin at temperatures above 15 C. may be due to several factors. A. 
maculatuin, in its natural environment, remains burrowed in moist leaf litter or 
rotten logs, coming to the surface to feed at night, and is most active on rainy 
nights. In this micro-environment, this species rarely encounters temperatures 
exceeding 20 C. It is possible, therefore, that during its evolutionary history, 
certain enzymes or other physiological systems became adapted to function optimally 
at temperatures approximating 15 C. to the point where they are not sufficiently 
labile to be altered significantly by changes in acclimation temperatures. If en- 
zyme systems of A. maculatuin are adjusted to function optimally at temperatures 
approximating 15 C., higher temperatures could result in decreased oxygen 
uptake through the skin. 

SUMMARY 

1. In Ambystoma maculatuin, the lungs and buccopharyngeal mucosa become 
increasingly important in respiration at higher temperature. 

2. The skin accounts for more than 50% of the total oxygen uptake at 15 C. 
and below. 

3. Approximately 80 % of the carbon dioxide produced is released through the 
skin at all temperatures except 5 C., where no measureable amount of carbon 
dioxide is released through the lungs and buccopharyngeal mucosa. 

4. Lung and buccopharyngeal tidal volumes increased directly with tempera- 
ture; and the rate of buccopharyngeal oscillations increased greatly at higher tem- 
peratures, while the rate of lung inspirations remained relatively constant. 

5. Buccopharyngeal oscillations are of appreciable importance in the respiration 
of A. maculatuin, especially at higher temperatures. 

LITERATURE CITED 

COLE, W. H., AND J. B. ALLISON, 1929. The pharyngeal breathing rate of the frog as related to 

temperature and other factors. /. Exp. Zool., 53: 411-420. 
CZOPEK, J., 1962. Vascularization of respiratory surfaces in some Caudata. Copeia, 1962: 

576-587. 
DOLK, H. E., AND N. POSTMA, 1927. liber die Haut und die Lungenatmung von Rana tcm- 

poraria. Zeitsclir. vergl. Physiol., 5: 417-444. 

ELKAN, E., 1955. The buccal and pharyngeal mucous membrane in urodeles. Proc. Zool. Soc. 
' London, 125 : 685-692. 



354 WALTER G. WHITFORD AND VICTOR H. HUTCHISON 

HUBBS, C. L., AND CLARK HUBBS, 1953. An improved graphical analysis and comparison of 

series of samples. Syst. Zool., 2 : 49-57. 
KROGH, A., 1904. On the cutaneous and pulmonary respiration of the frog. Skand. Arch. 

Physiol., 15:328-419. 
LAPICQUE, L., AND J. PETETIN, 1910. Sur la respiration d'un batracien urodele sans poumons, 

Euproctus montanus. C. R. Soc. Biol., 69: 84-86. 
MATTHES, E., 1927. Der Einfluss des Mediumwechsels aus das Geruchsvermogen von Triton. 

Zeitschr. vergl. Physiol., 5: 83-166. 
NOBLE, G. K., 1925. The integumentary, pulmonary and cardiac modifications correlated with 

increased cutaneous respiration in the Amphibia ; a solution of the "hairy frog" 

problem. /. Morph. Physiol., 40: 341-416. 
SCHOLTEN, J. M., 1942. A few remarks on the respiratory movements of the frog. Arch. Neerl. 

Set., 26: 250-268. 
Vos, H. L, 1936. tiber die Atembewegungen und den Schnuffelmechanismus (Kehloszillationen) 

bei Reptilein und Amphibien. Zool. Ans., 115: 142-144. 
WHIPPLE, I. L., 1906. The ypsiloid apparatus of urodeles. Biol. Bull., 10: 255-297. 




THE JUVENILE HORMONE. III. ITS ACCUMULATION AND 
STORAGE IN THE ABDOMENS OF CERTAIN 

MALE MOTHS 

CARROLL M. WILLIAMS 1 
Tin" Binliii/iciil Laboratories, Harvard University, Cambridge 38, Mtissucliusctt.-; 

In the previous papers of this series (Williams, 1959, 1961 ) a simple and 
highly selective test for the juvenile hormone was described in terms of its ability 
to suppress the transformation of pupae into adult moths. By the use of this so- 
called "pupal assay" the role of the juvenile hormone was investigated during 
successive stages of the metamorphosis of the Cecropia silkworm. 

A by-product of these studies was a muling which one could scarcely have 
predicted by any rational process ; namely, that the abdomens of male Cecropia and 
Cynthia moths contain a cache of juvenile hormone. The object of the present 
communication is to document this fact in experiments performed on five species 
of wild silkworms Cecropia, Cynthia, Polyphemus, Pernyi, and Orizaba. 

METHODS 

In addition to methods previously described (Williams, 1959, 1961), the fol- 
lowing special procedures were utilized : 

1. Pai'abiosis between pupae and headless moths 

The moth was deeply anesthetized with carbon dioxide. The antennae and the 
fore and midlegs were excised and melted wax was applied with a drawing-pen to 
cover the entire head and prothorax. Then, with a single transverse cut, the head 
was removed to leave a collar of wax at the anterior open end of the prothorax. 
Crystals of an equal-part mixture of phenylthiourea and streptomycin sulfate were 
placed in the wound along with enough Ringer's solution to fill the cavity. 

The pupal partner was deeply anesthetized with carbon dioxide and a disc of 
integument, about 4 mm. in diameter, was cut from the mesothoracic tergum. The 
underlying epidermis was trimmed away with microscissors, care being taken to 
avoid any damage to the aorta which extends beneath the midline at this point. 
Melted wax was applied to the integument around the margin of the wound. 
Crystals of phenylthiourea and streptomycin were placed in the wound and the 
cavity was filled with a few drops of Ringer's solution. 

The two animals were oriented in a cradle of plasticene and the wax-coated 
openings were brought into juxtaposition. The pupal abdomen was compressed 
until the pupal blood filled the narrow opening between the animals, all air being 
thereby displaced. The junction was then sealed with melted wax to yield a 
parabiotic preparation such as shown in Figure 1. 

1 This study was supported, in part, by a grant from the National Institutes of Health. 

355 



356 



CARROLL M. WILLIAMS 



The preparation was then removed from the anesthesia funnel and stored at 
25 C. Under this condition, the moth commonly initiated energetic flapping of 
its wings. In order to prohibit this activity, the wings were placed between the 
jaws of a spring-loaded clothes-pin. 

2. Parabiosis between f>uf>ae and adult abdomens 

Melted wax was applied to the first abdominal segment of an anesthetized moth. 
With a single transverse scissor-cut the entire abdomen was detached from the 
thorax. The wax-coated edges of the wound were spread apart and the glistening 
air-filled crop was grasped with forceps and removed. Crystals of the phenyl- 
thiourea-streptomycin mixture were placed in the wound along with a few drops 
of Ringer's solution. The adult abdomen was then joined in parabiosis with a 
pupal partner, as described above. A preparation of this type is shown in Figure 3. 

TABLE I 

Parabiosis between diapausing Cecropia pupae and headless moths 



Species of moth 


Sex of pupal 
partner 


Number of 
preparations 


Results 


Polyphemus 9 


9 


2 


Prolonged survival* but no 








development 


9 


c? 


2 


Prolonged survival but no 








development 


c? 


9 


2 


Prolonged survival but no 








development 


cT 


cf 


2 


Prolonged survival but no 








development 


Pernyi 9 


9 


2 


Prolonged survival but no 








development 


9 


cf 


2 


Prolonged survival but no 








development 


cf 


9 


2 


Prolonged survival but no 








development 


cf 


c? 


2 


Prolonged survival but no 








development 


Cecropia 9 


9 


2 


Prolonged survival but no 








development 


9 


rf 


2 


Prolonged survival but no 








development 


d 1 


9 


5 


Prolonged survival but no 








development (2) 








Pupae developed into moths 








retaining many pupal 








characters (3) 


c? 


rf 1 


4 


Prolonged survival but no 








development (3) 








Pupa developed into moth 








retaining many pupal 








characters ( 1 ) 



* Moths survived for up to 10 weeks ; pupae survived up to 6 months. 



JUVENILE HOKMONK 



357 



RESULTS 
1. Pafabiosis bctivcen diapausing Cecropia pupae and headless moths 

Thirty-five diapausing Cecropia pupae were joined in parabiosis with headle 
Polyphemus, Pernyi, and Cecropia moths. Six of the preparations died within 
a week at 25 C. and were discarded. The behavior of the 29 viable preparations is 
outlined in Table I. 

There are three points of interest in this table. The first is the spectacular pro- 
longation of life in moths joined to pupal partners. The second point of interest is 
the lack of any developmental response in the vast majority of preparations. But 
the most surprising finding of all is the fact that 4 of 9 diapausing pupae initiated 
development when joined to headless male Cecropia moths ; moreover, in each of 
these cases the pupa developed and molted into an adult which preserved numerous 

TABLE 1 1 
Parabiosis between previously chilled pupae and headless moths 



Species of moth 


Species of pupal 
partners 


Number of 
preparations 


Effects on pupal partner 


9 Polyphemus 


d" or 9 Polyphemus 


4 


Formed normal moths 


c? Polyphemus 


cf or 9 Polyphemus 


4 


Formed normal moths 




C? or 9 Cecropia 


4 


Formed normal moths 


9 Cecropia 


d" or 9 Polyphemus 


4 


Formed normal moths 




c? or 9 Cecropia 


4 


3 formed normal moths 








1 formed moth retaining a patch 








of pupal cuticle on thoracic 








tergum 


cf Cecropia 


cf or 9 Polyphemus 


2 


2 developed into moths retaining 








many pupal characters 




d" or 9 Cecropia 


8 


8 developed into moths retaining 








many pupal characters 



pupal characteristics throughout head, thorax, and abdomen. These four pupae, in 
short, behaved as if they had been implanted with active corpora allata (Williams 
1952a, 1959, 1961). 

2. Parabiosis between previously chilled pupae and headless moths 

The experiments now under consideration differed from the preceding in that 
the pupal partners possessed endocrinologically active brains and were therefore 
able to initiate adult development within a few days after being placed at room 
temperature. 

Forty preparations were assembled, of which ten soon died and were discarded. 
Table II summarizes the several types of experiments that were performed. In 
each of the 30 viable preparations the pupal partner initiated adult development 
within ten days at 25 C. Attention is directed to the effects of the parabiosis 
on the course of this development. 

When the headless partner was a male or female Polyphemus moth, the pupa 



358 



CARROLL M. WILLIAMS 




FIGURE 1. A headless male Cecropia moth is here joined in parabiosis with a previously 
chilled pupa of the Polyphemus silkworm. 

FIGURE 2. After three weeks at 25 C. the pupa has molted to form a second pupal stage 
showing only traces of adult characters. The old pupal cuticle has been removed. 

FIGURE 3. The abdomen of a male Cecropia moth is joined in parabiosis with a previously 
chilled Cecropia pupa. 

FIGURE 4. After three weeks at 25 C. the pupa has molted to form a second pupal stage 
showing only traces of adult characters. The old pupal cuticle has been removed. The adult 
abdomen has molted its adult cuticle which is here shown partially peeled away. 



Jl'VKNILE HOKMOXK 

underwent normal adult development. The same was true in 7 of <S preparations 
in which the headless partner was a female Cecropia moth. But in all ten prepara- 
tions in which the headless partner was a male Cecropia moth, the pupa meta- 
morphosed, not into a normal moth, but into a creature which retained large areas 
of pupal cuticle (Fig. 2). It seems necessary to conclude that a male Cecropia 
moth, though headless and without any corpora allata, can somehow favor the 
release of juvenile hormone within the parabiotic preparation. 

In the analysis of these experiments we have centered attention on the effects of 
the parabiosis on the pupal partner. But, what about the developmental response 
of the other half of the combination- the headless moth? For present purposes 
suffice it to say that in about a third of the preparations the developmental reaction 
of the pupa spread to the adult partner and caused the latter to molt. The molt 
extended over both the thorax and abdomen, but never included the wings. The 
old adult cuticle was detached from the underlying epidermis and replaced by a 
smooth new cuticle which was of adult type, except for the generalized absence 
of scales or hairs. 

3. Parabiosis between previously chilled pupae and moth thoraccs and abdomens 

Each of a series of ten male Cecropia moths was beheaded and subdivided into 
thorax and abdomen. The two parts were then joined in parabiosis with previously 
chilled Polyphemus pupae and placed at 25 C. 

Within the following month all the surviving pupae which had been joined 
to adult thoraces emerged as normal moths. By contrast, all of the surviving 
pupae which had been joined to adult abdomens developed into moths retaining 
prominent pupal characteristics. This result demonstrated that juvenile hormone 
activity was associated with the abdomen of male Cecropia moths and that the 
thorax was inert in this respect. 

4. Juvenile hormone activity in relation to the species and se.v of the moth abdomen 

Table III summarizes a large series of experiments which were performed to 
test the abdomens of male and female saturniid moths belonging to five different 
species. A total of 110 parabiotic preparations w T as established, of which 83 were 
viable. 

Here again we see that the abdomens of male Cecropia moths always provoked 
strong positive tests for juvenile hormone. The same result was obtained for the 
abdomens of male Cynthia moths. The only other species to give a positive test 
was one of six abdomens of male Orizaba moths. The abdomens of both male and 
female Polyphemus moths were inactive ; the same was true for the closely related 
Pernyi moth and for most males and all females of Orizaba. 

Juvenile hormone activity in these adult insects is therefore a species-specific 
characteristic which is seen most prominently in Cecropia and Cynthia moths. 
Even here, it is a sex-linked characteristic being routinely encountered in males and 
only rarely, if at all, in female moths. 

Though the outcome is strikingly dependent on the species and sex of the adult 
component in the parabiotic preparations, the result is seemingly independent of 
the species or sex of the pupal partner. This fact is emphasized in Table III 
where we see that male Cecropia abdomens gave uniformly positive tests when 



360 



CARROLL M. WILLIAMS 



TABLE III 

I'arubiosis between /jrevimtsly chilled 



and moth abdomens 



Adult abdomen 


Species of pupal 
partner 


Xu in her of 
preparations 


Effects on pupal partner 


9 Polyphemus 
cf Polyphemus 


cf or 9 Polyphemus 
cf or 9 Polyphemus 


4 
6 


Formed normal moths 
Formed normal moths 


9 Pernyi 
cf Pernyi 


cf or 9 Polyphemus 
cf or 9 Polyphemus 


2 

4 


Formed normal moths 
Formed normal moths 


9 Orizaba 
cf Orizaba 


cf or 9 Polyphemus 
cf or 9 Polyphemus 


3 

6 


Formed normal moths 
5 formed normal moths 
1 formed moth retaining a patch 
of pupal cuticle on thoracic 
tergum 


9 Cecropia 
cf Cecropia 


cf or 9 Polyphemus 
9 Cecropia 
cf or 9 Polyphemus 

9 Cecropia 
cf Cecropia 


3 
5 
21 

13 

4 


Formed normal moths 
Formed normal moths 
Developed into moths retaining 
many pupal characters 
Developed into moths retaining 
many pupal characters 
Developed into moths retaining 
many pupal characters 


9 Cynthia 
cf Cynthia 


cf or 9 Polyphemus 
cf or 9 Polyphemus 


6 
6 


Formed normal adults 
Developed into moths retaining 
many pupal characters 



joined to pupae of male Cecropia, female Cecropia, male Polyphemus, or female 
Polyphemus. 

In about a third of the preparations the adult abdomen underwent a molt in 
synchrony with the development of the pupal partner (Figs. 4 and 7). This 
phenomenon will be considered in further detail in the Discussion. 

5. Temporary parabiosis between previously chilled pupae and abdomens of male 
Cecropia moths 

How long must a pupa lie joined to a male Cecropia abdomen before it will 
give a positive test for juvenile hormone? This question was studied in six prepa- 
rations in which the initial parabiosis was disassembled after a certain time and the 
adult abdomen grafted to a fresh pupa. 

The results were as follows : When the initial parabiosis was for only one day, 
the pupa developed into a normal moth, whereas the second pupal partner developed 
into a mixture of pupa and adult. But when the first parabiosis was extended to 
five days, then both the first and the second pupal partners gave positive tests for 
juvenile hormone. A transient blood connection with the adult is therefore 
sufficient. 

In additional experiments testing this point it appears that the parabiosis must 
continue until the pupal partner actually initiates its adult development ; in the 
above-mentioned experiments this occurred on the fourth day. If the pupa re- 



JUVENILE HORMOXK 



361 



quires more than four davs to initiate its development, then the parabiosis must he 
continued until it does so. 

6. Effects of allatecloiny 

The results show that abdomens of male Cecropia and Cynthia moths are some- 
how able to cause a positive test for juvenile hormone when joined to pupal part- 
ners. Since the adult corpora allata are cephalic structures, the moth abdomen 
did not appear to be a reasonable candidate for the secretion of juvenile hormone. 
Attention therefore centered on the corpora allata in the pupal partner. Despite 
the fact that pupal corpora allata are known to be inactive CWilliams, 1961). it 
seemed possible that they might lie "turned on" by some influence arising in the 
moth abdomen. In order to test this possibility, the series of experiments, sum- 
marized in Table IV, was carried out. 

Corpora allata were excised from four previously chilled Cecropia pupae. 
Then, each allatectomized pupa was joined to a headless male Cecropia moth. 
Though the preparations now contained no corpora allata, a positive test for 
juvenile hormone was obtained in each case. This result leaves no reasonable 
doubt that the source of the juvenile hormone was the adult abdomen itself. This 
conclusion was tested as follows : 

The corpora allata and corpora cardiaca were excised from six chilled male 
Cecropia pupae which were then placed at 25 C. and allowed to develop into adult 
moths. Three of these allatectomized moths were beheaded and joined to pre- 
viously chilled pupae ; in the other three, the moth abdomens were excised and 
used in the parabiosis. As recorded in Table IV. all six preparations now gave 
negative tests for juvenile hormone. This shows that the accumulation of juvenile 
hormone by the abdomen is dependent on its synthesis and secretion by the corpora 
allata in the head. 

In a further test, the corpora allata and corpora cardiaca were excised from two 
previously chilled male Cecropia pupae and two pairs of "loose" pupal corpora 



TABLE IV 

Parabiosis between previnuslv dulled pupae and male Cecropia moths; cither the pupal or 

adult partner had been allatectomized 



Adult component 



Headless cf Cecropia 



Headless cf Cecropia (alla- 
tectomized in pupal stage) 

cf Cecropia abdomen (alla- 
tectomized in pupal stage) 

cf Cecropia abdomen (alla- 
tectomized in pupal stage 
and two pairs of pupal 
corpora allata re-implanted 
into head) 



Pupal partner 



Allatectomized 
9 Cecropia 

cf Cecropia 
cf Cecropia 
cf Cecropia 



Xumber of 
experiment? 



3 
3 

2 



Effects on pupal partner 



Developed into moths 
retaining many pupal 
characters 

Formed normal moths 

Formed normal moths 

Developed into moths 
retaining many pupal 
characters 



362 CARROLL M. WILLIAMS 

allata-cardiaca were re-implanted into the head. When the moths emerged, their 
abdomens were joined to chilled pupae. The latter gave positive pupal assays for 
juvenile hormone (Table IV). This indicates that implanted corpora allata can 
substitute for the animal's own corpora allata in the secretion of juvenile hormone. 

7. Iihition of juvenile hormone from abdominal tissues of male Cecropia moths 

On the basis of the experiments just considered, it seems necessary to con- 
clude that the juvenile hormone is secreted by the corpora allata in the head of the 
adult moth and that, in the case of male Cecropia and Cynthia moths, the abdomen 
is somehow able to bind and accumulate substantial amounts of hormone. We are 
led to the prediction that the abdomen of the male Cecropia moth must contain a 
depot of juvenile hormone. A direct test of this inference was made as follows : 

The abdomens of several male Cecropia moths were dissected in Ringer's solu- 
tion. Various tissues and organs were removed, rinsed, and tested by implantation 
under a plastic window at the tip of the abdomen of previously chilled pupae. The 
results, summarized in Table V, show that positive assays for juvenile hormone 

TABLE V 

Tests for juvenile hormone in abdominal tissues and organs of adult male Cecropia moths* 
Adult tissue Xuniber of preparations Results ot tests 



1 negative 

3 negative 

4 negative 



Fragment of fat-body 3 2 positive 

Abdominal ganglia and connectives 4 1 positive 

Vas deferens 5 1 positive 

Testes 2 Negative 

Gut 1 Negative 

Solution of egg yolk from 3 Negative 
adult female 

" The tissues or organs were implanted into the tip of the abdomen of previously chilled 
Cecropia or Polyphemus pupae. The test was scored as positive when the pupa formed a moth 
retaining pupal characters. 

were obtained in certain cases after the implantation of adult fat-body, abdominal 
nerve cords, or vas deferens. The positive test in each case was a minimal reaction 
in that pupal cuticle was re-formed only under the abdominal window at the site 
of injury (Williams, 1961). A number of other adult organs, including a solu- 
tion of yolk obtained from unfertilized Cecropia eggs, gave negative tests. 

DISCUSSION 

1. The sequestering oj juvenile hormone 

The experimental results reveal the surprising fact that the abdomens of male 
Cecropia and Cynthia moths contain a cache of juvenile hormone. The hormone is 
found in the most prominent tissue of the abdomen the fat-body and also in at 
least two other tissues, the nerve cord and the vas deferens. Moreover, the hor- 
mone can be eluted from these tissues when the latter are either implanted or joined 
to a test pupa by way of the hemolymph. This shows that the sequestered hormone 
is in some sort of dynamic equilibrium with the circulating blood. 



JUVENILE HORMONE 



363 



Though tlu 1 hormone accumulates in the abdomen, it is synthesized hv ihe 
corpora allata in the head ( Fig. 5). So, il the corpora allata are excised prior to 
the final week of adult development, then, as diagrammed in Figure 0. no hormone 
accumulates in the ahdomen. 

It will be recalled that the corpora allata are inactive in the pupa and that this 
inactivity persists during two-thirds of adult development (Williams, 1961 ). Then, 
on or ahout the fourteenth day of adult development, the corpora allata recover the 
activity they had lost months earlier at the time of pupation. Yet, strange to say, 
the juvenile hormone has no known function in the adults of these short-lived 
Lepidoptera. Allatectomized Cecropia pupae develop into normal male and female 
moths which, when cross-mated, give rise to fertile eggs (Williams, 1959). 

Despite persistent uncertainties as to the role of juvenile hormone in these adult 
insects, we can confidently say that the hormone is sequestered in the abdominal 




FIGURE 5. As indicated by the stippled area, the juvenile hormone which accumulates in 
the abdomen of the male Cecropia moth is synthesized and secreted by the corpora allata in 
the head. 

FIGURE 6. If the corpora allata are removed from the head, then no hormone accumulates 
in the abdomen. (Figures 5 and 6 are used with the permission of the Scientific American.) 

tissues of male Cecropia and Cynthia moths. This process presumably begins after 
the corpora allata regain their activity during the final week of adult development, 
and continues during the entire week of adult life (Williams, 1961). The ab- 
domen, in effect, serves as an extraction chamber which, over a period of about 
two weeks, binds and accumulates the hormone which is continuously secreted by 
the corpora allata. 

2. The species- and se.v-specific accumulation of juvenile hormone 

The accumulation of hormone takes place in male Cecropia and Cynthia moths, 
but only rarely in females. Yet the corpora allata of female moths are just as 
active in secreting juvenile hormone as are those of males (Williams, 1959). 
Moreover, the accumulation of juvenile hormone occurs in neither male nor female 



364 CARROLL M. WILLIAMS 

I'ernyi or Polyphemus moths despite the fact that the adult corpora allata of 
these insects are extremely active in the secretion of juvenile hormone (Williams, 
1959). 

For these several reasons it is amply evident that those moths which fail to 
accumulate the hormone must dispose of it in some manner, presumahly by con- 
verting it into inactive products. Whatever this inactivating mechanism may be, 
it is obviously curtailed in the case of male Cecropia and Cynthia moths. 

3. The induced ino/thif/ of adult moths 

Among the pterygote insects only the Ephemeroptera are known to molt as 
adults a phenomenon which has recently been subjected to detailed study by 
Taylor and Richards (1963). Yet, as demonstrated in the present study, one can 
often cause an adult moth to molt by joining it in parabiosis with a pupal partner. 

The molt consisted of a detachment and retraction of the epidermis from the 
adult cuticle and the secretion of a new cuticle ; in no case did the moth make any 
effort to escape from the exuviae. The molting of the adult began in synchrony 
with that of the pupal partner. When the latter was a diapausing pupa and failed 
to initiate development (Table I), then no trace of molting occurred in the adult. 
This shows that the molting of the moth was a response to the ecdyson secreted by 
the pupal prothoracic glands and conveyed to the moth in the pupal blood. 

Manifestly, the cells and tissues of the moth retain the potential for further de- 
velopment and molting when supplied with ecdyson. The failure of adult insects 
to molt is therefore directly attributable to the breakdown of their prothoracic 
glands and the consequent lack of ecdyson. 

In the parabiotic preparations molting occurred in only about one-third of the 
moths. This capricious result was apparently due to uncontrollable variations 
in the circulation of blood between the pupal and adult partners. The blood 
connection which one can establish at the anterior end of a headless moth is, 
perforce, of small diameter. And when an adult abdomen was used, the blood flow 
from the pupa was commonly impaired by the herniation into the pupa of the moth 
ovaries and the fluid-filled rectal sac. In many of the adult partners these cir- 
cumstances apparently precluded the build-up of a threshold titer of the ecdyson. 

In most preparations the molting of the moth was incomplete and one could 
not remove the loose cuticle because of its persistent attachments to the spiracles 
and genitalia. However, in occasional preparations the molt was complete and the 
old cuticle could be easily peeled from the abdomen. 

A preparation of this type is shown in Figure 7. The entire adult cuticle has 
been molted, including the intricate cuticle of the male genitalia. The new cuticle 
preserves the smooth texture and pale tan pigmentation of normal adult cuticle, the 
only striking difference being the virtual absence of scales or hairs. The naked 
character of the new adult cuticle was especially prominent in Cecropia and was 
apparently due to the depletion of "scale mother cells" which during the forma- 
tion of the first adult cuticle were transformed into lifeless scales and sockets 
(Henke, 1946). However, in the case of Polyphemus or Pernyi abdomens, the 
new adult cuticle commonly showed substantial clusters of scales along the dorsum 
of the abdominal segments ; this apparently signifies the presence in these species 
of a reserve supply of scale and socket "stem cells." 



Jl'YEXILK HORMONE 365 

hi all of the- preparations that molted, a new adult cuticle formed, irrespective 
of whether the molt occurred in the presence or ah.sence of juvenile hormone. The 
new cuticle showed no trace of the reappearance of pupal characteristics even when 
the molt took place in the presence of sufficient juvenile hormone to cause the pupal 
partner to form a second pupal stage. This result is therefore different from that 
reported for Rhodnius where, according to Wigglesworth (1940, 1 ( '5S), a partial 




FIGURE 7. The abdomen of a male Cecropia moth was joined with a previously chilled pupa 
of the Cecropia silkworm. Here, after four weeks at 25 C., the pupa has transformed into a 
moth, preserving numerous pupal characters. Meanwhile, the moth abdomen has fully molted its 
adult cuticle and formed a new cuticle of adult type, except for the absence of scales and hairs. 

''reversal of metamorphosis" occurs when an adult is caused to molt in the presence 
of juvenile hormone. 

4. Prolongation of adult life 

In all the species used in the present study the moths are unable to feed because 
of the absence of functional mouth parts. Therefore, they rarely survive at room 
temperature for more than a week. And yet, when beheaded and joined to dia- 
pausing pupae, these same moths routinely survived for many weeks, during which 
time one could elicit leg movements and the flapping of the wings. Even the most 
moribund moth, within a few hours of death, reacted to parabiosis with renewed 
vitality and longevity. 



366 CAKUOI.L M. WII.UAMS 

Manifestly, the moths arc sustained h\ the fluids and suhstratcs which arc con- 
tinuously transfused rw the pupal hlood. This shows that the moths normally 
die of desiccation and nutritional deficiencies, and that the adult cells and tissues 
are capable of a far longer life-span than they can ordinarily display. 

5. Sc.nial reflc.vcs oj headless moths and isolated abdomens 

Many species of insects show exaggerated sexual reflexes following decapitation 

an effect attributable to a release from the normal inhibitory influences of the 

cephalic ganglia (Roecler ct al., 1960). Effects of this sort were particularly 

prominent in parabiotic preparations where the decapitated individual survived for 

up to ten weeks. 

The effects were most spectacular in the headless males or isolated male abdo- 
mens of Pernyi and Cecropia. Here, immediately after recovery from anesthesia, 
the individual initiated lively motion of both its abdomen and claspers, precisely as 
in normal mating. This behavior usually continued for several days, and was 
accompanied by the discharge of one or more spermatophores from the extended 
aedeagus. This curious behavior on the part of the headless male moth or male 
abdomen was far less conspicuous in Orizaba and totally lacking in Polyphemus. 

In general, the effects of decapitation were not as prominent in the case of 
female moths. One generally observed a great increase in the tone of the inter- 
segmental muscles and a downward flexion of the tip of the abdomen as in normal 
egg-laying. However, no eggs were actually oviposited, even though the abdomen 
was full of them. 

6. The mimicking of brain hormone by juvenile hormone 

According to experiments reported previously (Williams, 1959), the implanta- 
tion of active corpora allata can cause the initiation of development in a certain 
proportion of brainless diapausing pupae. On further analysis, the conclusion was 
reached that under certain conditions the juvenile hormone can activate the pro- 
thoracic glands and in this sense mimic the function of the brain hormone. 

These earlier results are strikingly reminiscent of those recorded in Table I. 
Here, 4 of 9 diapausing Cecropia pupae were caused to initiate development by 
joining them in parabiosis with headless male Cecropia moths. The net effect in 
this case, as in the experiments reported four years ago, was the termination of 
diapause and the production of pupal-adult monstrosities. 

The stimulation of development in the pupal partner, as noted in Table I, was 
observed when diapausing pupae were joined to male Cecropia moths which, as we 
have seen, contain a rich depot of juvenile hormone. In the absence of sequestered 
hormone, headless female Cecropia moths, as well as headless Polyphemus and 
Pernyi of both sexes, gave negative tests in all cases. 

Consequently, the experiments summarized in Table I are additional evidence 
that the juvenile hormone can mimic the brain hormone in its ability to turn on 
the prothoracic glands. 

SUMMARY 

1. Pupae joined in parabiosis with headless male Cecropia moths behave as if 
they have received an injection of juvenile hormone. They develop, not into 



JUVENILE HORMOXK 367 

normal moths, but into creatures which show a mixture of pupal and adult 
characters. 

2. By diverse experiments it was possible to show that the juvenile hormone 
comes from the moth abdomen and that the abdominal tissues of male Cecropia 
moths contain a rich depot of juvenile hormone. 

3. In the moth, itself, the hormone is synthesized by the corpora allata in the 
head and is progressively bound and sequestered by the abdominal tissues. If the 
corpora allata are removed from the head, then no hormone accumulates in the 
abdomen. 

4. The accumulation of juvenile hormone in the abdominal tissues occurs in male 
Cecropia and Cynthia moths, but not in females. In the case of certain related 
species of saturniid moths (Polyphemus, Pernyi and Orizaba), neither sex is 
ordinarily able to accumulate the hormone, despite the fact that they have very 
active corpora allata. 

5. The failure to accumulate the hormone points to some unknown means for 
its inactivation ; these agencies are evidently curtailed or by-passed in the case of 
male Cecropia and Cynthia moths which accumulate large amounts of hormone. 

6. Adult moths are frequently caused to molt when joined to non-diapausing 
pupae and thereby supplied with ecdyson. A new adult cuticle forms which is 
deficient in scales and hairs. Adults molting in the presence of high concentrations 
of juvenile hormone show no reappearance of pupal characters or any sign of a 
"'reversal of metamorphosis." 

7 . "When the adult tissues are continuously perfused with pupal blood, the life- 
span of the moths is greatly prolonged. This shows that, in the absence of func- 
tional mouthparts, the moth normally dies of desiccation and starvation rather than 
from the intrinsic biological death of the tissues themselves. 

8. In experiments involving the parabiosis of diapausing pupae with moths 
containing a depot of juvenile hormone, additional evidence was obtained that 
juvenile hormone can turn on the prothoracic glands and, in this sense, mimic the 
brain hormone. 

LITERATURE CITED 

HEXKE, K., 1946. Ueber die verschiedenen Zellteilungsvorgange in der Entwicklung des 

beschuppten Fliigelepithels der Mehlmotte Ephestia kuhniclla. Biol. ZbL, 65: 120-135. 
ROEDER, K. D., L. TOZIAN AND E. A. WEiANT, 1960. Endogenous nerve activity and behavior 

in the mantis and cockroach. J. Insect Physio!., 4: 45-62. 
TAYLOR, R. L., AND A. G. RICHARDS, 1963. The subimaginal cuticle of a mayfly (Callihuctis sp. ). 

Ann. Eiit/nnol. Soc. Am. (in press). 
WIGGLESWORTH, V. B., 1940. The determination of characters at metamorphosis in Rhmlnius 

prolixus (Hemiptera). 7. Exp. Biol, 17: 201-222. 
WIGGLES WORTH, V. B., 1958. Some methods for assaying extracts of the juvenile hormone in 

insects. 7. Insect. Pliysiol.. 2: 73-84. 
WILLIAMS, C. M., 1952a. Morphogenesis and the metamorphosis of insects. The Hur:',-y 

Lectures. 47: 126-155. 
WILLIAMS, C. M., 1952b. Physiology of insect diapause. IV. The brain and prothoracic glands 

as an endocrine system in the Cecropia silkworm. Biol. Bui!., 103: 120-138. 
WILLIAMS, C. M., 1959. The juvenile hormone. I. Endocrine activity of the corpora allata of 

the adult Cecropia silkworm. Biol. Bull., 116: 323-338. 
WILLIAMS, C. M., 1961. The juvenile hormone. II. Its role in the endocrine control of molting, 

pupation, and adult development in the Cecropia silkworm. Biol. Bull., 121: 572-585. 



FORMATION OF ENDODERM FROM ECTODERM 
IN CORDYLOPHORA 

EDGAR ZWILLING 1 

Dcpt. of Biology, Brandeis University, Waltham 54, Mass. 

There has been no satisfactory resolution for the problem of the extent to 
which one type of differential animal cell may be converted to that of another type. 
In one guise or another this question has come into prominence a number of times 
during the past half century. Several decades ago the issue was whether "dedif- 
ferentiation" followed by "redifferentiation" in a new direction is possible. The 
observations of H. V. Wilson (1907, 191 la, 191 Ib) on reconstitution from dis- 
sociated sponges and hydroids stimulated one phase of interest in the problem. 
Contributions to the discussion were made by many outstanding biologists of that 
time (e.g., J. S. Huxley, C. M. Child, J. Loeb, T. H. Morgan, etc.). Later, ex- 
periments with cultured tissues led to the general observation that rapidly pro- 
liferating tissues lose their characteristic morphological appearance. This stimu- 
lated more research and thinking on the "dedifferentiation-redifferentiation" issue. 
The major aspects of the current attitudes in this field have been reviewed by 
Trinkaus (1956) and Grobstein (1959). 

In general there are two schools of thought : one, which maintains that cell types 
are irreversibly fixed and that one type of cell cannot be transformed to another 
( except within a rather narrow "moclulatory" range) and the other, which claims 
that at least some types of cell transformation of a more radical sort (metaplasia) 
are possible. A number of well substantiated cases of metaplasia have been de- 
scribed in recent years. Our knowledge of Wolffian regeneration (formation of a 
lens from the iris of the eye) has been expanded by L. S. Stone and collaborators 
(1952). Fell and Mellanby (1953) and Weiss and James (1955) have demon- 
strated that the ordinarily squamous epithelium from the skin of 6-day chick 
embryos ma}' be converted to a mucous secreting columnar epithelium by ex- 
posure to excess vitamin A. The tissues of the chorio-allantoic membrane of 
chick embryos may change in response to a variety of conditions (Moscona, 1959, 
1960; Moscona and Carneckas, 1959). Other cases may be cited. There also are 
well substantiated cases which indicate that some tissues and some cell types are 
quite stable and do not transform to other types. This is especially true of so- 
called terminal tissues, i.e., tissues which have achieved an extreme state of 
functional maturity (possibly involving eventual cell sloughing) or extreme cell 
specialization. Ordinarily such cells do not divide. There remains a broad area in 
which the picture is not clear and it seems as though the best approach to the prob- 
lem is to state it in some form as "which cell types in which tissues are stable and 
which are sufficiently unstable so that they may undergo metaplasia under desig- 

1 Supported by a grant (RG-6334) from the National Institutes of Health. 

368 



FORMATION OF ENDODERM. FROM ECTODERM 369 

nated circumstances?" and then to design experiments which may answer this 
question. 

A complicating factor for this general problem is the possible existence, in many 
organisms and tissues, of reserve cells with "embryonic" capabilities. In some- 
situations, as in the case of regeneration of amphibian limbs, the importance of re- 
serve cells has been discounted (Butler, 1935). In other cases (i.e., interstitial 
cells (I-cells) of hydroids, amebocytes of sponges and neoblasts of flat worms) the 
reserve cell notion still has wide acceptance (Brien and Reniers-Decoen, 1955; 
Tardent, 1960; Burnett, 1962). Because of this situation, answers to the ques- 
tions raised above may be equivocal unless the experiments are designed to deal 
with progenies of single cells of known origin (i.e., clones). Unfortunately, the 
technical aspects of cloning cells have not yet been perfected for most cells coming 
from a wide range of tissues and organisms. In fact, the application of cloning 
procedures is still quite restricted. Since this is the case our present approaches 
must be less direct and less decisive. This communication deals with an ex- 
periment of the latter sort. While not unequivocal, it makes a contribution to our 
understanding of cell capacities in hydroids. 

Three papers, dealing with reconstitution of hydroids, may be cited as typical 
of the attitudes one encounters in such studies. Gilchrist (1937) found that ecto- 
derm and endoderm could be separated quite readily in scyphistomae of Aiirclia. 
The mesoglea is sufficiently thick in these animals so that it may be cut with a knife, 
leaving ectoderm with one half and endoderm with the other. The endodermal 
sheets formed small ciliated balls which lived for days but did not reconstitute an 
individual. The ectoderm, on the other hand, did form an individual complete 
with an endodermal layer. Gilchrist felt that the new inner layer must have been 
formed by the interstitial cells. - 

Beadle and Booth (1938) were, inter alia, concerned with the question of 
whether I-cells, largely restricted to ectoderm in Cordylophora, can form endo- 
dermal elements. To this end they separated ectoderm from endoderm in re- 
constitution masses made from the coenosarc of Cordyloplwra. Four masses of 
supposedly pure ectoderm were isolated. One of these developed a small hydranth 
and sections of two of the others revealed that they had begun to form an inner 
layer. The authors were uncertain of the nature of the inner layer in these 
two cases, since the cells were not typically endoderm. However, they felt that 
the fully reconstituted hydranth came from a case in which (p. 312) "... a small 
amount of endoderm must have been included. . . ." These are the two typical 
"explanations" : I-cells formed the new tissue or some cells from the tissue in ques- 
tion "must have been included" ! To complete the cycle I may cite Papenfuss and 
Bokenham (1939) who were unable to obtain reconstitution with either isolated 
ectoderm or endoderm from Hydra. These authors suggest the possibility that 
Gilchrist did not remove all of the endoderm from his supposed ectodermal 
isolates ! 

-Gilchrist did not include a histological study of his material with his account. Steinberg 
(1%3) repeated Gilchrist's experiments with Aitrclia scyphystomae and, from her histological 
study, found that these animals do not have typical interstitial cells. She did confirm the finding 
that a new endodermal layer formed from isolated ectoderm and presented evidence that prolifer- 
ating ectodermal cells give rise to amoehocytic cells which then form a typical endodermal layer. 
The amoebocytes themselves do not divide actively ; instead, the somatic ectodermal cells show 
considerable mitotic activity during the time that amoebocytes increase in number. 



370 



EDGAR ZWILLING 




FORMATION OF ENDODKKM FROM ICCTODKKM .$71 

Beadle and Booth dealt with only four masses of supposedly endoderm-frrr 
ectoderm, and an inner layer of some sort was formed in three of these. The orig- 
inal question which concerned Beadle and Booth was whether I-cells of the ecto- 
derm can form endoderm. Their conclusion that an endodermal layer can form 
only when the ectoderm is contaminated with former endoderm cells and that 
I-cells in the ectodermal layer cannot form endodermal elements seems to he 
somewhat shaky in view of their scanty evidence. Before a judgment can he made 
on the chief problem about the developmental capacity of the ectodermal cells it is 
important to establish more firmly the answer to the first question : does an endo- 
dermal layer form only from pre-existing endodermal cells or. put the other way. 
can an endodermal layer form from uncontaminated ectoderm? This question lias 
been subjected to close scrutiny with the Cordylophora lacustris found near Woods 
Hole as the test object. The results do not support Beadle and Booth but, instead, 
indicate that an individual, complete wth an endodermal layer, may be reconstituted 
from isolated ectoderm of Cordylophora. 

MATERIAL AND METHODS 

The Cordylophora used for these experiments was collected during July and 
August of two successive summers at the edge of a fresh-water pond. The best 
spot for collection was near a wooden run-off at one end of the pond. A small 
clump of the mature colony was placed in a jar with some of the pond water, taken 
to the laboratory and kept cool and aerated in the original pond water. Only fresh 
material (i.e., in the laboratory no more than three days) was used in the 
experiments. 

While the Cordylophora lived in fresh water, it was found that a more favorable 
medium for reconstitution of masses of coenosarc was 25-30% sea water (Beadle 
and Booth, 1938, used 50% sea water). The sea water was filtered, diluted with 
either tap or distilled water and then pasteurized. The pasteurized medium was 
cooled and aerated before use. 

Healthy, relatively straight stems were selected. The distal hydranth was cut 
away, and the coenosarc was pushed out of the perisarc by means of a curved bit 
of thin glass rod or a bit of shaped Tygon tubing which was held in a pair of 
forceps. Usually the gleaming white ectoderm separated from the pigmented endo- 
derm (yellow or orange) very readily (Fig. 1). An exposure of the stem to the 
25% sea water for 5-30 minutes before the extrusion greatly facilitated the separa- 
tion of layers. 

FIGURE 1. Newly extruded coenosarc of Cordylophora. Note the separation of ectoderm 
and endoderm (darker tissue) in the upper portion of the preparation. 

FIGURE 2. Section through 10-hour mass of equal parts of ectoderm and endoderm. Note 
the complete separation of the two layers and the presence of a thin mesogleal layer between the 
two. Note also that cells are intact. 

FIGURE 3. High power view of same section as in Figure 2. Note the inclusions in the 
endothermal cells (at left). 

FIGUKK 4. Section through 10-hour mass of pure ectoderm. Note the thin outer wall and 
highly fragmented and disorganized contents of the hollowed sphere. Compare with Figure 2. 

FIGURE 5. High power view of same section as in Figure 4. Note the character of the cells 
of the wall ; also the interstitial cell adhering to its inner surface. 



EDGAR ZWILLING 

Beadle and Booth found that neutral red was taken up selectively by the endo- 
derm and took advantage of this to distinguish the two layers. Since the natural 
pigmentation of the endoderm in our animals was so distinctive and the layers 
separated quite readily, no supplementary treatment was deemed necessary. 
Separation was completed with the aid of cataract knives and the ectoderm was 
removed to a clean dish for further inspection. The tissue was cut into small 
pieces (0.1-0.2 mm. in diameter) and each piece was carefully inspected, at high 
magnifications of the dissecting microscope (50-90 X), for contaminating endo- 
derm. Any pigmented material was removed from the ectoderm. Reflected and 
transmitted as well as oblique lighting were used to insure the identification and 
elimination of the orange endodermal cells. These could be seen most favorably 
against a white background. In this way there was reasonable certainty that all 
endodermal cells were removed. Several of the small masses were then heaped 
together and allowed to fuse and form a single mass. Masses were transferred to 
fresh medium in small Stender dishes and kept on a shaded part of the laboratory 
table. 

Despite all of the precautions taken there was always the possibility that one or 
two endodermal cells could escape our scrutiny and be included inadvertently. In 
order to assess the consequences of this possibility we set up control masses of 
ectoderm to which minute traces (no more than 3-5 cells) of endoderm were de- 
liberately added. In addition other controls with varying amounts of endoderm 
(up to 50%) were used. 

The table-top temperature, which was recorded for each experiment, varied 
from 19 to 22 C. Ectodermal masses were fixed in Bouin's fixative at various 
times following isolation; these were embedded in paraffin and sectioned at 5-10 /A. 
Toluidine blue at pH 5 proved to be a favorable stain. 

RESULTS 

A total of 117 viable masses of Cordylophora coensarc was prepared for this 
study. Forty of these consisted of pure ectoderm (i.e., no endoderm discernible 
with high magnifications of the dissecting microscope), 40 had deliberately included 
traces of endoderm, and in the rest endoderm comprised up to 50% of the total 
tissue mass. Thirty-five masses were fixed at various times after isolation. 

Pieces of extruded coenosarc fused readily and, in most cases, formed a smooth 
sphere. In favorable cases the sphere became hollow within five to 10 hours after 
isolation. Spheres which did not become hollow did not reconstitute, but re- 
mained in this inactive condition for days. The center gradually became opaque 
but the surface layer still retained the retractile appearance usually associated with 
living tissue. The hollowing-out process was noted by Beadle and Booth, who 
conjectured about the activity of ectoderm in transporting fluid to the interior of 
the mass. 

I Almost 50% of the ''pure" ectoderm masses reconstituted an inner layer (as 
f judged by the presence of clearly discernible flagellar motion in the enteron) and 
most of these (9) formed a complete hyclranth in 5-9 days. The rest (4) re- 
mained as stolonic growths which persisted without change for a long time. Essen- 
tially the same behavior was noted in masses which had a minute trace of endoderm. 
In both of these classes of isolates many (circa 50%.) of the masses remained in- 



FORMATION OF ENDODERM FROM HCTODKKM 373 

active (see above), and did not reconstitute. In sharp contrast all but one of the 
50:50 masses reconstituted a hydranth (one formed a stolonic growth) ; none were 
inactive. 

Most of the larger 50:50 masses formed a hydranth very quickly, within 30-36 
hours. The average time for hydranth formation was 2-3 days. This is the 
average for all size classes ; it was noted, however, that larger masses reconstituted 
a hvdranth more quickly than smaller ones. When the masses are distributed into 
two groups, those larger and those smaller than 0.2 mm., the average time for hv- 
dranth formation for the group of larger isolates (6 masses) was 1.1 days and that 
for the smaller (13 masses) was 2.4 days. Moreover, the hydranths formed In- 
most of the smaller mass were incomplete and had fewer tentacles than normal. A 
number of the large masses, on the other hand, developed more than one hydranth. 
(See Chalkley, 1945, for relation between initial mass and morphogenesis in 
Hvdra. ) 

*s 

HlSTOLOGICAL OBSERVATIONS 

Our histological study was made on 25 masses (10 were lost or were not satis- 
factory for study ) which were fixed at various times during reconstitution. Twelve 
of these were masses of supposedly pure ectoderm, eight were masses which had a 
trace of endoderm, and five consisted of equal amounts of ectoderm and endoderm. 

One of the first questions which required an answer was whether an occasional 
endoderm cell could be identified in the midst of a mass of ectoderm. Examination 
of sections of whole stems revealed that most of the endoderm cells were full of 
large inclusions of various sorts. This was the situation except in the hydranth and 
in the region immediately proximal to it where inclusion-free endoderm cells may 
be found. However, these regions were cut away before coenosarc was expressed 
and, therefore, most of the endoderm with which we are concerned was character- 
ized by the presence of many inclusions. Ectoderm cells are not free of inclusion 
but rarely have more than one or two per cell. On the basis of this criterion we 
could ascertain that supposedly endoderm-free masses had no cells with more than 
1-2 inclusions while a few cells with numerous inclusions could be identified in 
masses deliberately contaminated with traces of endoderm. The combination of 
this observation with the ease of identification of even single orange cells (endo- 
derm ) in the living mass when high magnifications of the dissection microscope 
were used provides reasonable confidence that our masses were endoderm-free. 
I doubt that one can make a stronger statement in a situation of this sort. Neither 
criterion is absolute ; masses which are allowed to live can be, at the best, only 
good approximations of those sectioned. 

The histological picture derived from masses of equal amounts of ectoderm and 
endoderm is quite clear and simple. The two layers rapidly sort out and. by the 
/end of ten hours, form a definite inner endodermal layer which is separated from 
the ectoderm by a distinct basement-membrane-like mesoglea (Figs. 2 and 3). 
Cells of both layers are intact and retain the appearance which is typical of an 
intact stem. When the hydranth is elaborated many of the endoderm cells acquire 
the characteristics of typical secreting cells of the digestive portion of a hydranth. 
The two salient features are: cells remain intact, and sorting into two distinct layers 
is rapid. 



374 



EDGAR ZWILLING 







FORMATION OF ENDODKRM FROM H( "K )l >KRM 375 

Since there was no observable difference between masses of "pure" ectoderm 
and those with traces of endodenn these two groups shall he treated together. 
Masses in these categories are characterized by a dramatic phenomenon. Sections 
of isolates which were fixed 1-2 hours after preparation revealed a picture which 
was quite similar to that seen in the ectoderm of sectioned intact stems. Somatic 
ectoderm cells and a number of I-cells could be distinguished clearly. By 8-10 
hours all of the active masses were hollow and sections revealed that the relatively 

J 

thin outer wall consisted of a single layer of small ectodermal cells. In isolates 
which were fixed at 10 hours or somewhat later, the contents of the sphere included, 
for the most part, cell fragments and scattered I-cells and relatively few recog- 
nizable intact somatic cells (Figs. 4, 5, 6). In some specimens it could be seen 
that a few I-cells were closely adherent to the inner surface of the cells of the 
sphere's wall (Fig. 5). 

At this stage, it was evident that many small basophilic cells were scattered 
through the cell debris. Some of the somatic cell nuclei were pycnotic (this was 
confirmed, in another series, with iron hematoxylin-stained material) but there 
were not enough degenerating nuclei to account for the total number of initial 
somatic cells. On the other hand there was a marked increase in the number of 
small basophilic cells (I-cells) but no indication of mitotic activity. 

During subsequent days the picture was essentially the same as for the 10-hour 
masses except that more I-cells accumulated on the inner surface of the wall (Figs. 
7 and 8). By the fourth day a complete inner layer of enlarged cells was found 
in all masses which had shown signs of internal flagellar beating prior to fixation. 
Mitotic figures were not evident during the interval from 10 hours to 4 days. By 
the time a hydranth had formed (5-9 days), the inner layer was typically endo- 
dermal (Fig. 9). During the later phases of this process (4 days and following) 
the endodermal cells in many of the masses contained inclusions. The decrease in 
amount of cell debris in the hollow of the mass leads to the suspicion that the cells 
were phagocytizing the debris. 

It is difficult to draw definite conclusions from static evidence of the sort pre- 
sented above. There is the possibility that the picture reconstructed from this 
particular series is not typical. There is no certainty that the new endodermal 
layer forms from the interstitial cells which are seen adhering to the inner surface 
of the wall of the spheres. Surviving somatic ectodermal cells could, conceivably, 
make this inner layer. Efforts are under way to obtain more rigorous evidence 
which may shed light on these problems. However, one aspect of the original ques- 
tion seems reasonably clear : an endodermal layer does form from pure ectoderm. 
This point has been established more firmly by the following additional observations. 

FIGURE 6. High power view (oil) of fragmented cellular material in interior of 24-hour 
ectodermal isolate. Note at least one small basophilic cell, many enucleated cell fragments and 
inclusions, and occasional (upper left) intact somatic cell. This photograph is from the same 
section as Figure 7. 

FIGURE 7. Section of 24-hour ectodermal isolate. Note two groups of intensely basophilic 
cells adhering to the inner surface of the sphere's wall. 

FIGURE 8. High power view of one of two groups of basophilic cells shown in Figure 7. 
.Note the typical interstitial cell characteristics of these cells. 

FIGURE 9. Section through a fully formed hydranth which has formed from an ectodermal 
isolate. 



376 EDGAR ZW1LLING 

TWO-STEP OPKKATION 

Several jucts could he derived from the observations described above. Within 
the first ten hours after separation of a mass of ectoderm a hollow sphere was 
formed, the wall of the sphere consisted of small ectodermal cells, all contaminating 
inadvertent or deliberately included endoderm cells were in the center of the 
sphere and were fragmented (or fragmenting) along with the rest of the ectoderm. 
In addition to this there were no I-cells /";/ the ectodermal wall, but a few may 
have been adhering to its inner surface. An operation was carried out which could 
take advantage of the condition at 10 hours. Eleven ectodermal masses were pre- 
pared in the usual way. After 10-12 hours these were opened up and the inner 
contents were extruded and discarded. The inner surface of the walls was scraped 
clean of adherent material. The walls were then cut up and allowed to fuse to- 
gether to form 5 second-generation reconstitution masses. 

All of the second-generation masses became hollow and, by the sixth day after 
the second step of the operation, all five of them had an endodermal layer which 
could be detected by the definite flagellar movements which were readily discernible 
in the intact masses. Three of these elongated and formed stolonic growths while 
the other two formed typical tentacled hydranths by the end of the seventh day. 
This last experiment removes any doubt that an endodermal layer can form from 
isolated ectoderm of Cordylophora. 

DISCUSSION 

The observations presented above indicate that the answer to the question of 
whether an endodermal layer can form from ectoderm of Cordylophora is different 
than that arrived at by Beadle and Booth. On the basis of a careful scrutiny of 
the material I feel confident that endoderm-free preparations of Cordylophora can 
reconstitute an endodermal layer. In addition to the histological confirmation of 
the absence of endoderm there is the two-stage operation which, in my mind, 
provides the most rigorous evidence in support of this contention. 

Equally rigorous evidence about the ectodermal cell type which forms the 
endoderm is not at hand. Present observations do not allow us to ascertain 
whether only pre-existing interstitial cells or former somatic ectodermal cells con- 
tribute to the new layer. We hope to present additional evidence on this point in 
a subsequent publication. However, there is evidence from other forms which 
indicates that similar metaplastic changes may take place in the absence of I-cells. 
Normandin (1960) has reported (without details) that ectoderm-free endoderm of 
Hydra will reconstitute a new ectoderm when the endodermal isolate is free of 
I-cells. In my laboratory S. N. Steinberg (1963) has reinvestigated Gilchrist's 
observations with ectoderm from scyphistomae of Aurelia. She has confirmed the 
formation of an endodermal layer in the ectodermal isolates but her histological 
studies have failed to reveal an ectodermal cell type which has the characteristics 
of an I-cell. The nature of I-cells and their possible contribution to morphogenetic 
processes should, at the present, be regarded with a great many reservations. 

The loss of tissue integrity and the accompanying cellular fragmentation which 
are typical of the ectodermal isolates of Cordylophora are striking phenomena. 
They appear to be very similar to the "histolysis" of muscle cells which has been 
noted in regenerating urodele limbs (Butler, 1935; Thornton, 1938; Hay, 1959) 



FORMATION OF ENDODERM FROM ECTODERM 377 

where the muscle nuclei with sonic 1 of the cell cytoplasm pinch off and acquire the 
characteristics of regeneration cells while the enucleated ]>arts of the cell degenerate 
and are eventually removed by phagocytes. Cellular fragmentation of Cordyloplioni 
ectoderm seems to be related to an imbalance in the relative amounts of ectoderm 
and endoclerm. In several cases in which single small masses (0.1 mm. in diam- 
eter) of endoclerm were included with considerably larger amounts of ectoderm, 
tissue integrity was maintained locally, and a mesoglea was quickly re-established 
at the site of contact between endoclerm and ectoderm. Ectoderm adjacent to 
this region of normal tissue association underwent the usual fragmentation. It is 
interesting to speculate that the key to the fragmentation may reside in whether or 
not a mesoglea persists or is re-formed. Isolated ectoderm of Aiirelia scyphystomae 
does not undergo fragmentation ; the thick mesoglea retains its usual relation to the 
cells. There seems to be a relationship between the persistence of a mesoglea or 
its rapid re-establishment and retention of cell integrity. This may merely be a 
coincidence, but it is an intriguing one. 

SUMMARY 

Within ten hours after isolation of ectodermal masses of Cordylophora lacustris, 
a hollow sphere is formed in which the outer wall consists of a single layer of rela- 
tively small cells and most of the inner cells are fragmented and disorganized. Dur- 
ing the course of 5-9 days an entire hydroid forms complete with an endodermal 
layer. A two-stage operation, in which only the thin outer walls of 10-hour iso- 
lates were kept for further observation, also resulted in the formation of new 
endodermal layers. While it is not clear whether interstitial or somatic ectoderm 
cells are involved, it is evident that some ectodermal derivative forms the new 
layer. The possible relationship between cell fragmentation and absence of a 
mesoglea is noted. 

LITERATURE CITED 

BEADLE, L. C., AND F. A. BOOTH, 1938. The reorganization of tissue masses of Cordylophora 

lacustris and the effect of oral cone grafts, with supplementary observations on Obclia 

gclatinosa. J. Ex p. Biol, 15: 303-326. 
BRIEN, P., AND M. RENIERS-DECOEN, 1955. La signification des cellules interstitielles des hydres 

d'eau douce et le probleme de la reserve embryonnaire. Bull. Biol. France ct Belg., 89: 

258-325. 
BURNETT, A. L., 1962. The maintenance of form in Hydra. In: Regeneration. Twentieth 

Growth Symposium. Ronald Press, New York. Pp. 27-52. 
BUTLER, E. G., 1935. Studies on limb regeneration in X-rayed Amblystoma larvae. Anat. Rcc., 

62 : 295-307. 
CHALKLEY, H. W., 1945. Quantitative relation between the number of organized centers and 

tissue volume in regenerating masses of minced body sections of Hydra. J . Nat. Cancer 

lust., 6: 191-195. 
FELL, H. B., AND E. MELLANBY, 1953. Metaplasia produced in culture of chick ectoderm by high 

vitamin A. /. Physio!.. 119: 470-488. 
GILCHRIST, F. G., 1937. Budding and locomotion in the scyphistomas of Aiirelia. Biol. Bull.. 

72: 99-124. 
GROBSTEIN, C., 1959. Differentiation of Vertebrate Cells. In: The Cell. Braclu-t and Mirsky, 

Editors. Academic Press, New York. Pp. 437-496. 
HAY, E. D., 1959. Electron microscopic observations of muscle differentiation in regenerating 

Amblystoma limbs. Dcrcl. Biol., 1 : 555-585. 
MOSCONA, A., 1959. Scjuamous metaplasia and keratinization of chorionic epithelium of the 

chick embryo in egg and in culture. Dcvcl. Biol., 1 : 1-23. 



F.lHiAU /WILLING 

MOMOXA, A., 1960. Mctaphistic changes in the chorioallantoic membrane. Transpl. Bull., 26 
120-124. 

MOSCONA, A., AND Z. I. CAKNKCKAS, 1959. Etiology of keratogciiic metaplasia in the chorioal- 
lantoic membrane. Science, 129: 1743-1744. 

XnkMANDiN, D. K., 1960. Regeneration of Hydra from the endoclerm. Science, 132: 678. 

PAPENFUSS, E. J., AND N. A. H. BOKENHAM, 1939. The fate of the ectoderm and endoderm of 
Hydra when cultured independently. Biol. Bull., 76: 1-6. 

STEINBERG, SONIA N., 1963. The regeneration of whole polyps from ectodermal fragments of 
scyphistoma larvae of Aitrclia aitrita. Biol. Bull., 124: 337-343. 

STONE, L. S., 1952. An experimental study of the inhibition and release of lens regeneration 
in adult eyes of Triturus r. I'iridescens. J. Exp. Zool., 121 : 181-224. 

TARDENT, P., 1960. Principles governing the process of regeneration in hydroids. In: Develop- 
ing Cell Systems and their Controls. Eighteenth Growth Symposium. Ronald Press, 
New York. Pp. 21-43. 

THORNTON, C. R., 1938. The histogenesis of muscle in the regenerating forelimb of larval 
Amblystoma punctatuiu. J. Morph., 62: 17-47. 

TRINKAUS, J. P., 1956. The differentiation of tissue cells. Amcr. Nat., 90: 273-288. 

WEISS, P., AND R. JAMES, 1955. Skin metaplasia in vitro induced by brief exposure to vitamin 
A. Exp. Cell Res., Suppl. 3: 381-394. 

WILSOX, H. V., 1907. On some phenomena of coalescence and regeneration in sponges. /. E.\-p. 
Zool., 5: 245-258. 

WILSON, H. V., 191 la. On the behavior of dissociated cells in Hydroids, Alc\onaria, and 
Asterias. J. Exp. Zool., 11: 281-338. 

WILSON, H. V., 1911b. Development of sponges from dissociated tissue cells. Bull, (for 1910) 
U. S. Bur. Fisheries (Doc. 750), 30: 1-30. 



INDEX 






A BE, S. The effect of p-chloromercuri- 

benzoate on amoeboid movement, flagellar 

movement and gliding movement, 107. 
Acclimatization of rainbow trout to different 

salinities, 45. 
Accumulation of juvenile hormone in moths, 

355. 

Acid phosphatase changes in planarians, 285. 
Acid phosphatase in snail digestive gland, 211. 
Activity in chromatophorotropins of Balanus, 

254. 

Agrotis, echoes of ultrasonic pulses from, 200. 
Alga, reproduction and ecology of Concho- 

celis colony of, 268. 
Alkaline phosphatase changes in planarians, 

285. 

Allatectomy of moths, effects of, 355. 
Amaroucium, significance of caudal epidermis 

in metamorphosis of, 241. 
Amathes, echoes of ultrasonic pulses from, 

200. 

Ambystoma, gas exchange in, 344. 
Aminopeptidase in snail digestive gland, 211. 
Amoeboid movement, effect of PCMB on, 107. 
Amphibian induction, 125. 
Amphipod, lunar orientation of, 97. 
Amphipods, osmotic regulation in, 225. 
Amphipyra, echoes of ultrasonic pulses from, 

200. 

Annelid, pelagic larvae of, 163. 
Anomalies in x-irradiated hybrid mice, 303. 
Antibiotics, role of in culturing crab larval 

stages, 141. 

Artemia, genetics of, 17. 
Ascidia, biology of, 31 

Ascidian metamorphosis, significance of cau- 
dal epidermis in, 241. 

Astronomical orientation of sandhopper, 97. 
Aurelia, regeneration of larvae of, 337. 
AUSTIN, C. R. Fertilization in Pectinaria 

(= Cistenides), 115. 

Autoradiography of dogfish thyroids, 170. 
Azide, effect of on ascidian metamorphosis, 

241. 

r> ACTERIA, role of in sea urchin gut, as 
related to division of endocommensal cili- 
ates, 1. 

Balanvs, cyclic activity in, 254. 

Barnacle, cyclic activity in, 254. 



Barnacle, temperature <>f, 277. 

BAKTH, L. G., AND L. J. BAKTII. The relation 
between intensity of inductor and type of 
cellular differentiation of Rana presump- 
tive epidermis, 125. 

Bats, perception of flying moths by, 200. 

BECK, S. D. See D. G. R. McLEoo, 84. 

BEERS, C. D. Relation of feeding in the sea 
urchin Strongylocentrotus to division in 
some of its endocommensal ciiiates, 1. 

BELAMARICH, F. A. Biologically active pep- 
tides from the pericardial organs of the 
crab Cancer, 9. 

Bicarbonate ion, effect of on Rana embryonic 
induction, 125. 

Biggaria, division in, as related to feeding of 
sea urchin, 1. 

Biology of Ascidia, 31. 

Body temperatures of marine tropical animals, 
277. 

BOWEN, S. T. The genetics of Artemia. II., 
17. 

BOYD, C. M., AND M. W. JOHNSON. Vari- 
ations in the larval stages of a decapod 
crustacean, Pleuroncodes, 141. 

Breathing of Ambystoma, 344. 

Breeding season of Pectinaria, 115. 

Brine shrimp, genetics of, 17. 

BRUNS, S. B. Sec L. E. LACHAXCE, 65. 

(^ ADDO, proprioceptors in legs of, 262. 

Calcium, effect of on Rana embryonic induc- 
tion, 125. 

Calliphorid fly, oogenesis and radiosensitivity 
in, 65. 

Campaniform proprioceptors in phalangid legs, 
262. 

Cancer, peptides from pericardial organs of, 9. 

Carbon dioxide, release of by Ambystoma, 344. 

Cardio-excitor material from crab pericc.rdial 
organs, 9. 

Caudal epidermis, significance of in meta- 
morphosis of ascidians, 241. 

Cecropia, juvenile hormone in, 355. 

Cell division in endosperm, factors in, 193. 

Cell fragmentation in Cordylophora, 368. 

Cell granules in snail digestive gland, 211. 

Cellular differentiation of Rana presumptive 
epidermis, 125. 



379 



380 



INDEX 



( rnfral IKTVOUS system of llalamis, chroma- 
tophorotropins of, 254. 

Changes in phosphatases in ])lanarians, 285. 

CHEVERIE, J. C, AND W. G. LYNN. High 
temperature tolerance and thyroid activity 
in the teleost fish, Tanichthys, 153. 

Chilling, role of in termination of insect dia- 
pause, 84. 

Chloride exchanges in rainbow trout, 45. 

Chloride in hydroids, 322. 

Chromatography of crab pericardia! organs, 9. 

Chromatophore control in Upogebia, 24. 

Chromatophorotropin activity in Balanus, 254. 

Chromosome number of Pectinaria, 115. 

Chromosomes of irradiated Cochliomyia fe- 
males, 65. 

Ciliates, endocommensal, relation of division 
in to feeding of sea urchin, 1. 

Cirripede, chromatophorotropin activity in, 
254. 

Climate of tropical intertidal zone, 277. 

CLONEV, R. A. The significance of the caudal 
epidermis in ascidian metamorphosis, 241. 

Cobalt-60 irradiation of Cochliomyia females, 
65. 

Cochliomyia, radiosensitivity of, 65. 

Coelenterate, formation of endoderm from 
ectoderm in, 368. 

Coelenterate, regeneration in, 337. 

Coelenterates, ionic relations in, 322. 

Cold, effect of in termination of insect dia- 
pause, 84. 

Cold, role of in thyroid activity of dogfish em- 
bryo, 170. 

Colonies of Porphyra, reproduction of, 268. 

Compass orientation of pigeons, 311. 

Conchocelis of Porphyra, ecology and repro- 
duction of, 268. 

Congenital anomalies in x-irradiated hybrid 
mice, 303. 

Contraction, equatorial, as factor in endo- 
sperm cytokinesis, 193. 

Contraction of epidermis, role of in meta- 
morphosis of Amaroucium, 241. 

Cordylophora, formation of endoderm from 
ectoderm in, 368. 

Cordylophora, ionic relations in, 322. 

Corn borer, photoperiodic termination of dia- 
pause in, 84. 

Corpora allata of moths, role of in production 
of juvenile hormone, 355. 

COSTLOW, J. D., JR. Molting and cyclic ac- 
tivity in chromatophorotropins of the cen- 
tral nervous system of the barnacle, 
Balanus, 254. 

Crab, peptides from pericardial organs of, 9. 

Crab, variations in larval stages of, 141. 



Crustacean, chromatophore control and nruro- 
secretion in, 24. 

Crustacean, eye-color mutant in, 17. 

Crustacean, peptides from pericardial organs 
of, 9. 

Crustacean, variations in larval stages of, 141. 

Crustaceans, water regulation in, 225. 

Culture methods for galatheid crab larvae, 
141. 

Culture of Porphyra Conchocelis, 268. 

Cutaneous and pulmonary gas exchange in 
Ambystoma, 344. 

Cyanide, effect of on ascidian metamorphosis, 
241. 

Cyclic activity in Balanus, 254. 

Cyclic factors in lunar orientation of sand- 
hopper, 97. 

Cynthia moths, juvenile hormone in, 355. 

Cysteine, effect of on cell movement, 107. 

Cytochemistry of snail digestive gland, 211. 

Cytokinesis, endosperm, factors in, 193. 

Cytology of Cochliomyia ovarioles, 65. 

Cytology of fertilization in Pectinaria, 115. 

IT^ARKNESS, role of in lunar orientation 
of sandhoppers, 97. 

Dark, role of in termination of diapause in 
insect, 84. 

Day-length, effect of on Porphyra Conchocelis, 
268. 

Day-length, role of in termination of insect 
diapause, 84. 

DEAN, D., AND P. S. HATFIELD. Pelagic lar- 
vae of Nerinides, 163. 

Decapod crustacean, variations in larval stages 
of, 141. 

Dedifferentiation in regenerating Aurelia lar- 
vae, 337. 

Developing dogfish, thyroid activity of, 170. 

Developing Rana, embryonic induction in, 125. 

Developing sea urchin eggs, resistance of to 
osmotic stress, 182. 

Development of Pectinaria, 115. 

Development of sea urchin egg, effect of 
lithium and IBA on, 55. 

Development of young ascidians, 31. 

Developmental stages of Cochliomyia ovari- 
oles, irradiation of, 65. 

Developmental stages of Galleria, nerve cord 
shortening during, 293. 

Diapause, photoperiodic termination of, in in- 
sect, 84. 

Differentiation of Rana presumptive epidermis, 
125. 

Digestive gland cells of snail, histochemistry 
of, 211. 



INDEX 



381 



DI.MOND, Su. M. T. The relation of whole- 
hod}' I 131 uptake to thyroid activity in the 
developing dogfish, Scyliorhinus, 170. 

Dipteran, oogenesis and radiosensitivity in, 65. 

Directional training of homing pigeons, 311. 

Displaced pigeons, sun compass orientation of, 
311. 

DITZION, B. Sec R. M. ROSENBAUM, 211. 

Division of endocommensal ciliates, relation of 
to feeding of sea urchin, 1. 

Dogfish embryos, thyroid activity of, 170. 

Duration of life in galatheid crab, 141. 

"p CDYSIS and cyclic activity in Balanus, 

254. 

Ecdysis in Galleria, 293. 
Echinoderm, effect of IBA and lithium on egg 

of, 55. 
Echinoderm, relation of feeding in, to division 

of endocommensal ciliates, 1. 
Echinoderm, resistance of to osmotic stress, 

182. 
Echoes of ultrasonic pulses from flying moths, 

200. 

Ecology of ascidians, 31. 
Ecology of Porphyra Conchocelis, 268. 
Ectoderm, formation of endoderm from, in 

Cordylophora, 368. 

Ectodermal fragments of Aurelia, regenera- 
tion of, 337. 
EDGAR, A. L. Proprioceptors in the legs of 

phalangids, 262. 
Effect of lithium on development of sea urchin 

egg, 55. 

Effect of PCMB on cell motion, 107. 
Egg, sea urchin, effect of lithium and IBA on 

development of, 55. 
Egg, annelid, fertilization in, 115. 
Eggs, sea urchin, resistance of to osmotic 

stress, 182. 
Electrophysiology of phalangid propriocept- 

ors, 262. 

Embryo, Rana, induction in, 125. 
Embryology of ascidians, 31. 
Embryology of Pectinaria, 115. 
Embryology of sea urchin, effect of lithium 

and IBA on, 55. 

Embryos, dogfish, thyroid activity of, 170. 
Embryos, sea urchin, resistance of to osmotic 

stress, 182. 

Enargia, echoes of ultrasonic pulses from, 200. 
Endocommensal ciliates of sea urchin, division 

in, as correlated with feeding of host, 1. 
Endocrine activity in barnacles, 254. 
Endocrine activity in moths, 355. 
Endoderm, formation of from ectoderm in 

Cordylophora, 368. 



Endosperm cytokinesis, factors in, 1<>3. 

Ennomos, echoes of ultrasonic pulses from 
200. 

Entodiscus, division in, as related to feeding 
of sea urchin, 1. 

Environmental temperatures of tropical ma- 
rine animals, 277. 

Enzyme activities in planarians, 285. 

Enzymic histochemistry of snail diue^tive 
cells, 211. 

Epidermis, caudal, significance of in ascidian 
metamorphosis, 241. 

Epidermis, presumptive, of Rana, differenti- 
ation of, 125. 

Equatorial contraction as a factor in endo- 
sperm cytokinesis, 193. 

Equatorially displaced pigeons, sun compass 
orientation of, 311. 

Esterases in snail digestive gland, 211. 

Exchange of gas in salamander, 344. 

Exchanges, chloride, in rainbow trout, 45. 

Eye-color in Artemia, mutation of, 17. 

p ACTORS of equatorial contraction and 

polar membrane expansion in endosperm 

cytokinesis, 193. 
FARMANFARMAIAN, A. See A. C. GIESE, 

182. 
Feeding, relation of to changes in phospha- 

tases in planarians, 285. 
Feeding in sea urchin, relation of to division 

of its endocommensal ciliates, 1. 
Female reproductive system, radiosensitivity 

of, in Cochliomyia, 65. 
Fertilization in Cistenides, 115. 
Fertilization in Pectinaria, 115. 
Fiddler crab as test object for barnacle 

chromatophorotropins, 254. 
FIXGERMAX, M., AND C. OcuRO. Chromato- 

phore control and neurosecretion in the 

mud shrimp, Upogebia, 24. 
Fish, osmotic relations of, 45. 
Fish, temperature tolerance and thyroid ac- 
tivity of, 153. 

Fissurella, temperatures of, 277. 
Flagellar movement, effect of PCMB on, 107. 
Fluorescence microscopy of Pectinaria eggs, 

115. 
Flying moths, echoes of ultrasonic pulses 

from, 200. 
Formation of endoderm from ectoderm in 

Cordylophora, 368. 
Fresh-water gammarids, osmotic regulation 

in, 225. 

/~J ALATHEID crab, variations in larval 

stages of, 141. 
Galleria, metamorphosis of nerve cord in, 293. 



382 



INDEX 



( i;tmm;i irradiation of Cochliomyia females, 
65. 

Gammarids, osmotic regulation in, 225. 

Gas exchange in salamander, 344. 

Gastropod, temperatures of, 277. 

Genetic effects in heterosis of x-irradiated hy- 
brid mice, 303. 

Genetics of Artemia, 17. 

Germination of Porphyra Conchocelis, 268. 

GIESE, A. C., AND A. FARMANFARMAIAN. 
Resistance of the purple sea urchin to os- 
motic stress, 182. 

Gliding movement of cells, effect of PCMB 
on, 107. 

Glucuronidase in snail digestive gland, 211. 

GOODBODY, I. The biology of Ascidia. II., 31. 

GORDON, M. S. Chloride exchanges in rain- 
bow trout adapted to different salinities, 
45. 

Granular components in digestive cells of 
snail, 211. 

Graptolitha, echoes of ultrasonic pulses from, 
200. 

Growth of ovaries in irradiated Cochliomyia 
females, 65. 

"LJ AGSTR6M, B. E. The effect of lithium 
and o-iodosobenzoic acid on the early de- 
velopment of the sea urchin egg, 55. 

HATFIELD, P. S. See D. DEAN, 163. 

Heat, role of in termination of insect dia- 
pause, 84. 

Heat, effect of on intertidal tropical animals, 
277. 

Heat tolerance and thyroid activity of fish, 
153. 

Helix, histochemistry of digestive gland of, 
211. 

Hemerocallis endosperm, cytokinesis of, 193. 

Heterosis in x-irradiated hybrid mice, 303. 

Hexapod nervous system, studies on, 293. 

High temperature, role of in termination of 
insect diapause, 84. 

High temperature tolerance and thyroid ac- 
tivity of fish, 153. 

High temperatures, effect of on intertidal 
tropical animals, 277. 

Histochemistry of planarians, 285. 

Histochemistry of snail digestive cells, 211. 

Histology of dogfish embryo thyroid, 170. 

Histology of fish thyroid, 153. 

Histology of metamorphosis in ascidians, 241. 
1 listology of regenerating Aurelia larvae, 337. 

Homing pigeons, sun compass orientation of, 
311. 

Hormonal control of molting and cyclic ac- 
tivity in barnacle, 254. 
1 lormone, juvenile, of moths, 355. 



HUTCHISON, V. H. Sec W. G. WHITKORD, 
344. 

Hybrid mice, x-ray-induced congenital anom- 
alies in, 303. 

Hydra, ionic relations in, 322. 

Hydranth regeneration in Aurelia, 337. 

Hydroid, formation of endoderm from ecto- 
derm in, 368. 

Hydroid larvae, regeneration of, 337. 

Hydroids, ionic relations in, 322. 

T BA, effect of on development of sea urchin 
egg, 55. 

In vitro culture of Porphyra Conchocelis, 268. 

Insect diapause, photoperiodic termination of, 
84. 

Intensity of inductor, role of in differentiation 
of Rana presumptive epidermis, 125. 

Intertidal tropical animals, temperatures of, 
277. 

o-Iodosobenzoic acid, effect of on development 
of sea urchin egg, 55. 

Ionic relations in hydroids, 322. 

Ionic relations in rainbow trout, 45. 

Ionizing radiation, effects of on Cochliomyia, 
65. 

IWASAKI, H., AND C. MATSUDAiRA. Observa- 
tions on the ecology and reproduction of 
free-living Conchocelis of Porphyra, 268. 

J OHNSON, M. W. See C. M. BOYD, 141. 
Juvenile hormone of moths, 355. 

If CN, effect of on ascidian metamorphosis, 
k 241. 

J^ACHANCE, L. E., AND S. B. BRUNS. 
Oogenesis and radiosensitivity in Cochli- 
omyia, 65. 

Larvae of ascidians, metamorphosis of, 241. 

Larvae of Aurelia, regeneration of, 337. 

Larvae of Nerinides, 163. 

Larval development of ascidian, 31. 

Larval stages of Pleuroncodes, variations in, 
141. 

Latitudinally displaced pigeons, sun compass 
orientation of, 311. 

Leiobunum, proprioceptors in legs of, 262. 

Lepidopteran, metamorphosis of nerve cord in, 
293. 

Lethal temperatures of fish, 153. 

LEWIS, J. B. Environmental and tissue tem- 
peratures of some tropical intertidal ma- 
rine animals, 277. 

Life-cycle of Porphyra, 268. 

Light, effect of on Porphyra Conchocelis, 268. 



INDEX 



383 



Light, role of in termination of insect dia- 
pause, 84. 

Lily endosperm cytokinesis, factors in, 193. 

Limpet, temperatures of, 277. 

Lithium, effect of on Rana embryonic induc- 
tion, 125. 

Low temperature, in relation to thyroid ac- 
tivity of dogfish embryo, 170. 

Low temperature, role of in termination of 
insect diapause, 84. 

Lunar orientation of sandhoppers, 97. 

Lungs, role of in respiration of Ambystoma, 
344. 

LYNN, W. G. Sec J. C. CIIKVHKIK, 153. 

\TADSEXIA, division in, as related to feed- 
ing of sea urchin, 1. 

Magnesium, effect of on Rana embryonic in- 
duction, 125. 

Male moths, accumulation and storage of 
juvenile hormone in, 355. 

Marine animals, temperatures of, 277. 

Marine gammarids, osmotic regulation in, 225. 

Marinogammarus, osmotic regulation in, 225. 

MATSUDAIRA, C. Sec H. IWASAKI, 268. 

Maturation of Pectinaria eggs, 115. 

McLEOD, D. G. R., AND S. D. BECK. Photo- 
periodic termination of diapause in an in- 
sect, 84. 

Mechanics of cytokinesis, study of, 193. 

Membrane expansion, polar, ac factor in endo- 
sperm cytokinesis, 193. 

Metamorphosis of ascidians, significance of 
caudal epidermis in, 241. 

Metamorphosis of nerve cord in Galleria, 293. 

Mice, hybrid, x-ray-induced congenital anom- 
alies in, 303. 

MILLER, A. T., JR. Sec P. J. OSBORNE, 285. 

Molting and cyclic activity in Balanus, 254. 

Morphology of crab larval stages, 141. 

Morphology of metamorphosis in ascidians, 
241. 

Moths, accumulation and storage of juvenile 
hormone in, 355. 

Moths, flying, echoes of ultrasonic pulses 
from, 200. 

Movement of cells, effect of PCMB on, 107. 

Mud shrimp, chromatophore control and 
neurosecretion in, 24. 

Muscle cells, significance of in metamorphosis 
of Amaroucium, 241. 

Alutation of eye-color in Artemia, 17. 

N AYKiATTON of pigeons, 31 1. 

Xephridium, role of in osmotic regulation in 

gammarids, 225. 
Nerinides, pelagic larvae of, 163. 



Xerita, temperature of, 277. 

Nervous system, hexapod, studies on, 293. 

Xervous system anomalies in x-irradialcd 

hybrid mice, 303. 
X'ervous system of barnacle, chromatophoro- 

tropins in, 254. 
Neurohormone from pericardia! organs of 

crab, 9. 

Neurometamorphosis in (ialleria, 293. 
Neurosecretion in Upogebia, 24. 
Neurosecretory activity in Balanus, 254. 
Xitzschia, effect on PCMB on movement of, 

107. 
Nutrition of planarians, histochemistry of, 285. 

QGURO, C. See M. FINGERMAN, 24. 

Oogenesis and radiosensitivity in Cochliomyia, 
65. 

Of-ilio, proprioceptors in legs of, 262. 

Orientation, lunar, of sandhoppers, 97. 

Orientation of pigeons, 311. 

Orizaba moth, juvenile hormone in, 355. 

Orthosia, echoes of ultrasonic pulses from, 
200. 

OSBORNE, P. J., AND A. T. MILLER, JR. Acid 
and alkaline phosphatase changes associ- 
ated with feeding, starvation and regen- 
eration in planarians, 285. 

Oscillatoria, effect of PCMB on movement of, 
107. 

Osmotic regulation of marine and fresh-water 
gammarids, 225. 

Osmotic relations of fish, 45. 

Osmotic stress, resistance of purple sea urchin 
to, 182. 

Ostrinia, photoperiodic termination of dia- 
pause in, 84. 

Ova, Pectinaria, fertilization in, 115. 

Ova, sea urchin, effect of lithium and IB A on 
development of, 55. 

Ovarian growth, effect of radiation on, in 
Cochliomyia, 65. 

Oxygen consumption of Amb.v-.toma, 344. 

pANDORIXA coenobia, effect of PCMB on 
flagellar movement of, 107. 

PAPI, F., AND L. PARDI. On the lunar orien- 
tation of sandhoppers, 97. 

Parabiosis in moths, 355. 

para-Chloromercuribenzoate, effect of on cell 
movement, 107. 

Parasite, oogenesis and radiosensitivity of, (>5. 

PARDI, L. Sec F. PAIM, >7. 

Pelagic larvae of Nerinides, 163. 

Peptides from era!) pericardia! organs, 9. 

Perchlorate, effect of on thyroid activity of 
dogfish embryo, I/O. 



384 



INDEX 



Pericardia! organs of crab, peptides from, 9. 

Permeability of rainbow trout gills, 45. 

Pernyi moth, juvenile hormone in, 355. 

Phalangid legs, proprioceptors in, 262. 

Phosphatase, acid, in snail digestive gland, 
211. 

Phosphatases in planarians, 285. 

Phospholipid in snail digestive gland, 211. 

Photoperiod, effect of on Porphyra Concho- 
celis, 268. 

Photoperiodic termination of diapause in in- 
sect, 84. 

Pigeons, sun compass orientation of, 311. 

Pigment control in Upogebia, 24. 

Pigment-dispersing substance of Balanus, 254. 

Pigment formation in Ascidia, 31. 

PIPA, R. L. Studies on the hexapod nervous 
system. VI., 293. 

Planarians, phosphatases in, 285. 

Pleuroncodes, variations in larval stages of, 
141. 

Polar membrane expansion as factor in endo- 
sperm cytokinesis, 193. 

Polychaete annelid, pelagic larvae of, 163. 

Polyphemus, juvenile hormone in, 355. 

Polyps of Aurelia, regeneration of, 337. 

Population dynamics of ascidians, 31. 

Porphyra, ecology and reproduction of Con- 
chocelis of, 268. 

Potassium in hydroids, 322. 

Potentials, action, from phalangid legs, 262. 

Prolongation of adult life in moths, 355. 

Proprioceptors in legs of phalangids, 262. 

Protozoa, endocommensal, relation of division 
in to feeding of sea urchin, 1. 

Pulmonary and cutaneous gas exchange in 
Ambystoma, 344. 

Pupae, moth, juvenile hormone in, 355. 

Pupae of Galleria, nerve cord shortening in, 
293. 

Pupation, termination of in insects, by photo- 
period, 84. 

Purple sea urchin, resistance of to osmotic 
stress, 182. 

Pyrallidae, metamorphosis of nerve cord in, 
293. 

RADIOCHLORINE, use of in study of 
osmotic relations of rainbow trout, 45. 

Radio-iodine uptake, relation of to thyroid ac- 
tivity of developing dogfish, 170. 

Radioresistance in hybrid mice, 303. 

Radiosensitivity of Cochliomyia, 65. 

Rainbow trout, chloride exchanges in, 45. 

Rana presumptive epidermis, differentiation 
of, 125. 

Regenerating planarians, phosphatases in, 285. 

Regeneration of Aurelia polyps, 337. 



Regulation, osmotic, of marine and fresh- 
water gammarids, 225. 

Relation between intensity of inductor and 
type of cellular differentiation of Rana 
presumptive epidermis, 125. 

Relation of feeding of Strongylocentrotus to 
division in its endocommensal ciliates, 1. 

Relation of whole-body radio-iodine uptake to 
thyroid activity in the developing dog- 
fish, Scyliorhinus, 170. 

Reproduction of Porphyra Conchocelis, 268. 

Resistance of purple sea urchin to osmotic 
stress, 182. 

Resorption of tail in ascidian larvae at meta- 
morphosis, 241. 

Respiration of Ambystoma, 344. 

Rhythm affecting lunar orientation of sand- 
hoppers, 97. 

Rhythm of chromatophorotropin activity in 
Balanus, 254. 

ROBERTS, H. S. Factors of equatorial con- 
traction and polar membrane expansion in 
endosperm cytokinesis, 193. 

ROEDER, K. D. Echoes of ultrasonic pulses 
from flying moths, 200. 

Roentgen irradiation of hybrid mice, 303. 

ROSENBAUM, R. M., AND B. DlTZION. Ell- 

zymic histochemistry of granular com- 
ponents in digestive gland cells of the 
Roman snail. Helix, 211. 

Rosette body of Upogebia, 24. 

RUGH, R., AND M. WOHLFROMM. X-irradi- 
ation-induced congenital anomalies in hy- 
brid mice, 303. 

gALAMANDER, gas exchange in, 344. 

Salinity relations of rainbow trout, 45. 

Salinity relations of sea urchin, 182. 

Salmo, chloride exchanges in, 45. 

Salt balance of rainbow trout, 45. 

Sandhoppers, lunar orientation of, 97. 

ScHMiDT-KoENir., K. Sun compass orienta- 
tion of pigeons upon equatorial and trans- 
equatorial displacement, 311. 

Screw-worm fly, oogenesis and radiosensitiv- 
ity in, 65. 

Scyliorhinus, thyroid activity in embryos of, 
170. 

Scyphistoma larvae of Aurelia, regeneration 
of, 337. 

Sea urchin, relation of feeding in to division 
in its endocommensal ciliates, 1. 

Sea urchin, resistance of to osmotic stress, 
1S2. 

Sea urchin egg, effect of lithium and II>A on 
development of, 55. 

Secretory-resorption cells of snail digestive 
gland, role of in digestion, 211. 



INDEX 



385 



Sensillae on legs of phalangids, 262. 
Sequential induction in Rana embryos, 125. 
Sex-linked mutation in Artemia, 17. 
Sexual development of irradiated Cochlomyia 

females, 65. 

Sexual reflexes of headless moths, 355. 
Shortening of ventral nerve cord in Galleria, 

293. 
Shrimp, mud, chromatophore control and 

neurosecretion in, 24. 
Significance of caudal epidermis in ascidian 

metamorphosis, 241. 
Skin, role of in respiration of Ambystoma, 

344. 
Slit organ proprioceptors in legs of pha- 

langids, 262. 
Snail, histochemistrv of digestive gland of, 

211. 

Sodium in hydroids, 322. 
Sperm penetration in Pectinaria, 115. 
Sporangia of Porphyra Conchiocelis, 268. 
Starvation, relation of to changes in phos- 

phatases in planarians, 285. 
Starvation of sea urchin, relation of to divi- 
sion of endocommensal ciliates, 1. 
STEINBACH, H. B. Sodium, potassium and 

chloride in selected hydroids, 322. 
STEINBERG, S. N. The regeneration of whole 

polyps from ectodermal fragments of 

scyphistoma larvae of Aurelia, 337. 
Sterility of Cochliomyia females, induced by 

irradiation, 65. 

Storage of juvenile hormone in moths, 355. 
Strongylocentrotus, relation of feeding in to 

division of its endocommensal ciliates, 1. 
Strongylocentrotus, resistance of to osmotic 

stress, 182. 

Studies on hexapod nervous system, 293. 
Sulfhydryl groups, importance of in cell move- 
ment, 107. 

Sun compass orientation of pigeons, 311. 
Sunira, echoes of ultrasonic pulses from, 200. 
Survival of young ascidians, 31. 

HP AIL epidermis of Amaroucium, role of in 
metamorphosis, 241. 

Tanichthys, temperature tolerance and thyroid 
activity of, 153. 

Teleost, osmotic relations of, 45. 

Teleost, temperature tolerance and thyroid ac- 
tivity of, 153. 

Temperature, role of in larval stage variation 
in Pleuroncodes, 141. 

Temperature, role of in photoperiodic termi- 
nation of diapause in insect, 84. 

Temperature, role of in respiration of Amby- 
stoma, 344. 

Temperature in relation to thyroid activity of 
dogfish embryo, 1/0. 



Temperature tolerance and thyroid activity of 

fish, 153. 

Temperatures of tropical marine animals, 277. 
Termination of diapause in insect, <S4. 
Tetraclita, temperature of, 277. 
Thiourea, effect of on fish thyroid, 153. 
Thiourea, effect of on thyroid activity of dog- 

fish embryo, 170. 
Thyroid activity of developing dogfish, in re- 

lation to whole-body I 131 uptake, 170. 
Thyroid activity and temperature tolerance of 

fish, 153. 

Tissue culture of Cordylophora, 368. 
Tissue temperatures of tropical marine ani- 

mals, 277. 
Tolerance, temperature, and thyroid activity 

of fish, 153. 
Trans-equatorially displaced pigeons, sun 

compass orientation of, 311. 
Tropical marine animals, temperatures of, 277. 
Trout, rainbow, chloride exchanges in, 45. 
Tubularia, ionic relations in, 322. 

T T CA as test object for barnacle chroma- 

tophorotropins, 254. 
Ultrasonic pulses, echoes of from flying moths, 

200. 
Upogebia, chromatophore control and neuro- 

secretion in, 24. 
Urinary rates of gammarids, 225. 

VARIATIONS in larval stages of Pleuron- 

codes, 141. 
Ventral nerve cord shortening in Galleria, 293. 

\A/"ATER balance of rainbow trout, 45. 

Water balance of sea urchins, 182. 

Water relations of gammarids, 225. 

Weight losses in intertidal tropical animals, 

277. 
WERNTZ, H. O. Osmotic regulation in ma- 

rine and fresh-water gammarids, 225. 
Whitecloud, temperature tolerance and thy- 

roid activity of, 153. 

WlIITFORD, W. G., AND V. H. HUTCHISOX. 

Cutaneous and pulmonary gas exchange 

in the spotted salamander, Ambystoma, 

344. 
Whole-body I i:!1 uptake and thyroid activity 

in developing dogfish, Scyliorhinus, 170. 
WILLIAMS, C. M. The juvenile hormone. III., 

355. 
WOHLFROMM, M. Sec 1\. Ri'cii, 303. 

V-IKKAD1. \TIOX 1 \IHVKI > anomalies 



'/WILLING, E. Formation of endoderm 
from ectoderm in Cordylophora, 368. 



MBL WHOI LIBRARY 




UH 1AZU