I>
•I
f 1
• f .
§.
- •. -::,-•
THE
BIOLOGICAL BULLETIN
PUBLISHED BY
THE MARINE BIOLOGICAL LABORATORY
Editorial Board
GARY N. CALKINS, Columbia University E. E. JUST, Howard University
E. G. CONKLIN, Princeton University FRANK R. LlLLIE, University of Chicago
E. N. HARVEY, Princeton University CARL R. MOORE, University of Chicago
SELIG HECHT, Columbia University GEORGE T. MOORE, Missouri Botanical Garden
LEIGH HOADLEY, Harvard University T. H. MORGAN, California Institute of Technology
M. H. JACOBS, University of Pennsylvania G. H. PARKER, Harvard University
H. S. JENNINGS, Johns Hopkins University EDMUND B. WILSON, Columbia University
ALFRED C. REDFIELD, Harvard University Managing Editor
VOLUME LXXIV
FEBRUARY TO JUNE, 1938
Printed and Issued by
LANCASTER PRESS, Inc.
PRINCE & LEMON STS.
LANCASTER, PA.
11
THE BIOLOGICAL BULLETIN is issued six times a year. Single numbers, $1.75. Subscription per volume (3 numbers), $4.50.
Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain: \Yheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W.C. 2.
Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Mass., between June 1 and October 1 and to the Institute of Biology, Divinity Avenue, Cambridge, Mass., during the re- mainder of the year.
ICntercd October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894.
LANCASTER I'Kl.SS. INC.. LANCASTER. PA.
CONTENTS
No. 1. FEBRUARY, 1938
PAGE
CLEVELAND, L. R.
Longitudinal and Transverse Division in Two Closely Related Flagellates 1
CLEVELAND, L. R.
Origin and Development of the Achromatic Figure 41
WAKSMAN, SELMAN A., AND UNTO VARTIOVAARA
The Adsorption of Bacteria by Marine Bottom 56
COE, W. R.
Primary Sexual Phases in the Oviparous Oyster (Ostrea vir- ginica) 64
SONNEBORN, T. M.
The Delayed Occurrence and Total Omission of Endomixis
in Selected Lines of Paramecium aurelia 76
REDFIELD, ALFRED C., AND ANCEL B. KEYS
The Distribution of Ammonia in the Waters of the Gulf of Maine 83
BRACKET, JEAN
The Oxygen Consumption of Artificially Activated and Fer- tilized Chaetopterus Eggs
TYLER, ALBERT, AND N. H. HOROWITZ
On the Energetics of Differentiation. VII 99
BARNES, T. CUNLIFFE
Experiments on Ligia in Bermuda. V. Further effects of salts and of heavy sea water 108
SUMMERS, F. M.
Some Aspects of Normal Development in the Colonial Ciliate Zoothamnium alternans 117
SUMMERS, F. M.
Form Regulation in Zoothamnium alternans
No. 2. APRIL, 1938 EARTH, L. G.
Quantitative Studies of the Factors Governing the Rate of
Regeneration in Tubularia
iii
jv CONTENTS
PAGE
KIDDER, GEORGE \Y., AND C. LLOYD CLAFF
Cytological Investigations of Colpoda cucullus. 178
NOVIKOFF, ALEX B.
Embryonic Determination in the Annelid, Sabellaria vulgaris.
I 198
NOVIKOFF, ALEX B.
Embryonic Determination in the Annelid, Sabellaria vulgaris.
II 211 PIERSON, BERNICE F.
The Relation of Mortality after Endomixis to the Prior Inter- endomictic Interval in Paramecium aurelia 235
GELBER, JULIUS
The Effect of Shorter than Normal Interendomictic Intervals
on Mortality after Endomixis in Paramecium aurelia. . . . 244
SPEICHER, KATHRYN G., AND B. R. SPEICHER
Diploids from Unfertilized Eggs in Habrobracon 247
MAXWELL, JANE
Inactivation of Sperm by X-Radiation in Habrobracon . . . 253
SCHMIEDER, RUDOLF G.
The Sex Ratio in Melittobia chalybii Ashmead, Gameto- genesis and Cleavage in Females and in Haploid Males (Hy- menoptera: Chalcidoidea) 256
HAYES, FREDERICK RONALD
The Relation of Fat Changes to the General Chemical Em- bryology of the Sea Urchin 267
ABRAMOWITZ, A. A., AND R. K. ABRAMOWITZ
On the Specificity and Related Properties of the Crustacean Chromatophorotropic Hormone 278
MAST, S. ()., AND COLEEN 1 < > \VLER
The Effect of Sodium, Potassium and Calcium Ions on Change- in Volume of Amoeba proteus. . . . 297
MAST, S. O.
The Contractile Yacunlr in Amoeba proteus (Leidy) 306
I.II.I.KLAND, OLE
Duration of Life without Food in Drosophila pseudoobscura . 314
YOUNG, K. T.
The Life History of a Trem.it ode (Levinseniella cruzi?) from the Shore Birds (Limos.i fecloa and Catoptrophorus semi- palmatus inornatus) 319
GIESK, ARTHUR C.
The Effects of I'ltra-Yiolet Radiation of X 2537A upon Cleav- age of Sea- Urchin Eggs 330
CONTENTS v
PAGE
OLSON, MAGNUS
The Histology of the Retractor Muscles of Thyone briareus Lesueur 342
No. 3. JUNE, 1938
OWEN, BRADFORD B.
A Simple Teleost Kidney in the Genus Cyclothone 349
WELSH, J. H., AND F. A. CHACE, JR.
Eyes of Deep-Sea Crustaceans. II. Sergestidae 364
LEAVITT, BENJAMIN B.
The Quantitative Vertical Distribution of Macrozooplankton
in the Atlantic Ocean Basin 376
MORGAN, T. H.
A Reconsideration of the Evidence Concerning a Dorso- ventral Pre-organization of the Egg of Chaetopterus 395
MORGAN, T. H., AND ALBERT TYLER
The Relation between Entrance Point of the Spermatozoon and Bilaterality of the Egg of Chaetopterus 401
SUMNER, F. B., AND P. DOUDOROFF
Some Experiments upon Temperature Acclimatization and Respiratory Metabolism in Fishes 403
STILES, KARL A.
Intermediate-winged Aphids and the Time-of-Determination Theory 430
GALTSOFF, PAUL S.
Physiology of Reproduction of Ostrea virginica, I 461
INDEX 487
Vol. LXXIV, No. 1 February, 1938
THE
BIOLOGICAL BULLETIN
PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY
LONGITUDINAL AND TRANSVERSE DIVISION IN TWO CLOSELY RELATED FLAGELLATES
L. R. CLEVELAND
(From the Biological Laboratories, Harvard University, Cambridge, Massachusetts}
The two flagellates considered in this paper are closely related morphologically, and are probably of common origin; yet one divides longitudinally in a typical flagellate manner, and the other trans- versely, as in ciliates. In the latter, the anterior daughter cell retains the parent extranuclear organelles, while the other daughter develops new organelles at the posterior end. This organism may represent the beginning stage in the development of a new group of flagellates.
The chromosomes of both organisms are interesting because they are large, two in number (except in rare instances), clearly differenti- ated in size, persist through the interphase, show structural details plainly, divide in the telophase, and are unmistakably moved to the poles by the extranuclear chromosomal fibres.
The achromatic figure is also interesting because the fibres com- posing it are as distinct as the flagella which the organisms possess, and it is plain that they have the same individuality as the flagella. They arise from the centrioles, which follow the flagellar bands, and are sometimes so intermingled with the flagella that they can be distinguished from them only by their function in nuclear division.
Eleven species of Spirotrichonympha have been described from termites but, with the exception of 5. polygyra, the species described by Cupp (1930), the descriptions are so brief that it is impossible to determine how many species are valid. And the description of Cupp left unanswered many vital questions concerning the structure and behavior of extranuclear organelles and the processes of nuclear division. Further, an examination of the protozoa of Kalotermes (Paraneotermes Light) simplicicornis (Banks), the termite harboring the S. polygyra described by Cupp, shows that her description dealt with two distinct organisms, and is therefore not only inadequate but, in many respects, inaccurate. However, it is not surprising that Cupp overlooked the fact that she was dealing with two organisms, because
1
L. R. CLEVELAND
they are so much alike in some respects — and yet so different in others. If the number of flagellar bands is not considered, their differences may be seen only when the processes of cell division are studied. These processes Cupp states she did not understand because of failure to find a sufficient number of successive stages.
An examination of the interphase flagellates of this termite shows that the number of organisms with two spiral flagellar bands is about equal to the number with four spiral flagellar bands. This fact Cupp noted, but she supposed that those with two bands were recently divided four-banded forms which would soon develop two new bands. But the number of two-banded forms is just as great in preparations without dividing organisms as in those with them. This alone indicates that the two-banded forms are not young or developing four-banded forms.
There are also forms with three, five, and six spiral flagellar bands, but these are very rare in comparison with the two- and four-banded forms. In all hypermastigote flagellates occasional irregularities occur in the flagellated areas, and therefore one should expect to find a few organisms, when countless thousands are studied, whose spiral flagellar bands likewise deviate from the normal numbers. These abnormalities are probably due, at least in part, as unpublished observations on B ar bid any mpha show, to exceptional behavior of the centriole, the organelle from which the flagellated areas arise; sometimes, one centriole fails to produce another from itself during cell division, in which case one daughter cell gets only one centriole while the other gets two; and at other times, one centriole produces two new centrioles instead of one from itself, thus giving one daughter three centrioles and the other two.
Organisms with one flagellar band also occur, and in preparations containing many dividing individuals they are present in greater numbers than the forms with three, five, and six flagellar bands; some of these are probably the result of irregularities in the behavior of the centriole, but most of them, as will be made clear presently, are posterior daughters produced in the transverse division of the body of the forms with two spiral flagellar bands.
Kven though Cupp considered the forms with two spiral flagellar to be identical with those having four bands, her species, • Polygyra, is valid for the forms with four bands because she con- sidered four to IK- the normal number of bands. It is only necessary to give a name to the form with two bands. This, however, is not an task, for its relationship to previously described species and genera must be considered from three angles: should it be placed in
LONGITUDINAL AND TRANSVERSE DIVISION
any of the three genera which possess two flagellar bands; should Spirotrichonympha, which possesses four flagellar bands, be amended so as to receive it; or should a new genus be erected for it. Except for the fact that it has two instead of four flagellar bands, it is indis- tinguishable from 5. polygyra when only interphase extranuclear organelles are considered, and its interphase nucleus does not differ greatly from that of S. polygyra; the four chromosomal coils are in two sheaths, while in 5". polygyra each of the four chromosomal coils is in a separate sheath. This difference results from the failure of the sheaths containing the daughter chromosomal coils to separate in the late telophase of the two-banded form. But when the processes of cell division, particularly the behavior of the centrioles, the direction of the achromatic figure, and the plane of cytoplasmic division, are considered the two-banded form is so different from S. polygyra that one wonders whether it is desirable to place it in the same genus with S. polygyra. On the basis of the chromosomes alone, no one familiar with hypermastigotes, especially those with spiral flagellar bands, would hesitate to separate S. polygyra generically from Macrospiro- nympha, Leptospironympha, and Spirotrichosoma, the previously described forms with two spiral flagellar bands, for S. polygyra has two rod-shaped chromosomes, while Macros pironympha, Leptospironympha, and Spirotrichosoma each has a fairly large number of V-shaped chromosomes. But the two-banded form under consideration also has two rod-shaped chromosomes which, like those of S. polygyra, are clearly differentiated morphologically into a short one and a long one; and it, too, should be separated from the previously described genera with two spiral flagellar bands. It is plainly a question, then, of whether one should separate S. polygyra and the two-banded form generically on a basis of four clear-cut differences; number of flagellar bands, behavior of centrioles, direction plane of the achromatic figure, and the plane of cytoplasmic division. Aside from the number of flagellar bands, the differences are largely physiological and the present tendency is not to place organisms in separate genera on this basis. This leaves for consideration only the number of flagellar bands. If the extranuclear organelles other than the spiral flagellar bands of S. polygyra and the two-banded form were at all different, the two organisms should be separated generically, but it has been impossible, after using various fixatives and stains, to detect a single difference. Hence, I believe the two organisms are more closely related than the number of flagellar bands indicates. In other words, the number of flagellar bands in this case does not represent a fundamental difference, and since it doesn't, it should not be used to separate the two organisms generically.
LIBRARY
teV ~ h
4 L. R. CLEVELAND
In view of the facts presented in the discussion above, the hyper- mastigote flagellate with two spiral flagellar bands in Kalotermes (Paraneotermes Light) simplicicornis (Banks) is placed in the genus Spirotrichonympha Grassi and Foa, 1911 and the species bisfnra, the species being new. In order to conserve space and to bring out more clearly the similarities and differences between S. polygyra and S. bispira, the morphology and processes of cell division in the two species are described in one paper.
I am indebted to Professor S. F. Light for a generous supply of the termites used in this study, to Miss Jane Collier for technical and research assistance, to Miss Dorothy G. Harris for making the drawings, and to the Penrose Fund of the American Philosophical Society for financial assistance. It was necessary to make and study eleven hundred permanent preparations before obtaining the infor- mation presented.
SPIROTRICHONYMPHA POLYGYRA
Morphology
In ten individuals the body ranged in length from 63 to 112 microns with an average of 81; the width at the widest portion ranged from 25 to 60 microns with an average of 45; the distance of the greatest width from the anterior end averaged 56 microns; the distance from the anterior end to the nucleus averaged 21 microns; the transverse diameter of the nucleus averaged 11 microns; the distance from the posterior termination of the spiral flagellar bands to the posterior end of the body averaged 26 microns; the length of that portion of the body covered by the flagellar bands (the flagellated area) averaged 55 microns; the number of turns or spirals made by the flagellar bands averaged 45; the length of flagella from the surface of the body averaged 20 microns; the length of flagella from the basal granules to the surface of the body averaged 7 microns; the total length of flagella averaged 27 microns; the distance from the flagellar bands (straight line) to the surface of the body averaged 5 microns, and the width of a flagellar band is about 1 micron.
In the anterior end the spirals made by the flagellar bands are small and close together and do not vary much for a distance of 5 to 7 microns. Then they gradually become larger and farther apart until the mid-region of the body is reached. From this point on to their termination there is little variation in their diameter and in tin- distance between them (Fig. 1). They turn sharply at their point of origin in the anterior end, and this gives them a ring-like appearance when viewed vertically. \Yhen viewed laterally at their point of
LONGITUDINAL AND TKANSN ICRSE DIVISION 5
origin, they may be seen to originate in two groups, two in each group (Text-fig. A, 3). In other words, there are two points from which the bands grow out posteriorly, and two of the bands arising from each point, lie side by side for a short distance. Sometimes, just before, but usually shortly after, cytoplasmic division the process of duplicating the flagellar bands begins. It cannot begin earlier, as in
JJHJflPS ^
r^rr.
TEXT-FIG. .1. Details of portions of Spirotrichonympha polygyra. 1. Surface view of spiral flagellar bands. X 3000. 2. Flat surface of a band, showing rows of basal granules and the flagella that arise from the granules. X 4800. 3. Anterior end, showing how the flagellar bands begin to spiral. X 4800.
some related genera, because the four bands of the parent cell must unwind first and distribute two bands to each daughter, a process that is not completed in this flagellate until just before cytoplasmic division (Figs. 28-37). If the two new or daughter bands were to grow out before the unwinding of the parent bands was completed or nearly so, they would become entangled with the parent bands and
6 L. K. CLEVELAND
could never lake their places besides them to form the iiiterphasc organism with four spiral llagellar bands.
When viewed laterally, depressions may be seen between the band- as shown in Text-fig. .-1 , 1. In other word-, the bands appear to be on ridges This impression is probably produced by the ertopla-mir layer, which extends Irom the surface of the body to the bands, pushing inward between the bands. The llai surface of the llagcllar bands does not lie parallel \\ith the antero-posterior axis of the body as do the bands in Macros f>iron yniplni, but lies at an angle, the po- terior margin of the band being nearer the surface of the body than the anterior, so that it is only when the bands are flattened that views of them as shown in Text-fig. .1. 2 may be obtained.
The basal granules, from which the llagella arise, lie in rows antero-posteriorly directed across the flagellar bands. \\ it h two granule- in each row (Text-fig. A, 2). The two llagella that arise from the granules in each row adhere to each other from their point of origin until they reach the surface of the body big. 1, Text-fig. .1, 1. 2). After leaving tin- band, the llagella extend posteriorly almost to the next band, from which point they turn someuhat abruptly and extend to the surface of the body, slanting posteriorly as shown in Fig. 1.
The large single axostyle is a hyaloid structure containing many fibres, extends from the anterior end through the central portion of the body, becomes wider as it approaches the nucleus which it sur- rounds, then smaller, enlarges again near the posterior end of the body, and after leaving the body tapers to a line point. In the portion anterior to the nucleus the axostylar fibres are grouped close together, so that a few ol them sometimes retain a small amount of stain in individuals suitably destained for nuclear details. This portion was mistaken for a centriole by ( upp (1930). By varying the detaining l inie, all degrees of des tain ing of t he axostyle max be obtained, ranging from a heavily stained structure to one \\ilh no stain at all.
The so-called parabasals differ greatly in appearance from those of most genera of flagellates, but their reaction to fixatives and stains is the same, and I presume they should be regarded as parabasals. Instead of being long and lender, they are spherical and tollo\\ the spirals of the flagellar bands from a point near the posterior border of the nucleus to the termination ol the bands 'I'ig. 1). They are probably attached to the band- since they follou them -o regularly and since they retain their relationship to the band- \\hen the band- and the body .ire broken up by drastic treatment; but it has been impossible to see a connection between them and the bands-.
There an- other strut line-, o\ei looked by ('upp and oilier- \\lio have studied Spirolrichonympfia, \\hich follou the llagellar bands just
LONGITUDINAL AND TRANSVERSE DIVISION 7
as closely as the parabasals and which are clearly connected to the bands by fine threads, each structure being connected by a single thread. These structures are considerably smaller than the parabasals, stain deeply following fixatives which do not render the parabasals stainable, and follow the bands a considerable distance anteriorly beyond the termination of the parabasals (Fig. 1). In every indi- vidual, they terminate at practically the same point, namely halfway between the mid-portion of the nucleus and the anterior end; and they extend posteriorly to the termination of the bands. But, unlike the parabasals, they lie closer together along the bands in the region of the nucleus and anterior to the nucleus than elsewhere. They, like the parabasals, probably arise from the bands, since various stages in their growth may be noted in new bands which are increasing in length posteriorly.
Mitosis
The chromosomes are duplicated in the late telophase (Fig. 15), and as a result. of this each interphase nucleus contains four chromo- somes (Figs. 3-7). Each chromosome contains a distinct coil within a sheath and is anchored to the anterior margin of the nuclear mem- brane by an intranuclear chromosomal fibre. These ends of the chromosomes are referred to as the anterior ends or as the attachment ends, while the opposite ends, which are attached either to a chromatin nucleolus or to portions of it, are referred to as the posterior ends. The portions of the chromatin nucleolus attached to the chromosomes vary considerably in size and shape (Figs. 3, 5) ; sometimes a single piece may be attached to all the chromosomes (Fig. 4) ; at other times a piece may be attached to each pair of chromosomes (Fig. 8). The daughter chromosomes are usually twisted around each other, the degree of twisting varying considerably in different nuclei. When all the chromosomes are attached to one nucleolus, they are twisted little if any, and in such nuclei the direction of the chromosomal coils may be followed plainly from one end of the chromosomes to the other (Fig. 4). It is also possible in many of the other nuclei where the chromosomes are each attached to a portion of the chromatin nucleolus and hence are permitted to twist considerably, to follow the coils clearly from the anterior to the posterior end of the chromosomes and to see that the coil sometimes reverses its direction in an individual chromosome (Fig. 3) ; and wThen each pair of daughter chromosomes is attached to a portion of the chromatin nucleolus, as illustrated in Fig. 8, the direction of the coils may be determined at a glance because they are so plain. A coil may reverse its direction at any point in a chromosome, and the coils of daughter chromosomes may be directed
8 L. R. CLEVELAND
either alike or differently. Models of chromosomal coils, made of wires, show that the shifting of direction of coiling in a chromosome does not interfere with the coil splitting lengthwise, as it does in the telophase, and separating to become two individual coils of daughter chromosomes.
The first step in cell division is the separation of the spiral flagellar bands at their anterior ends, two of the bands moving in one direction and two in the other; and as they separate they unwind, a process that is not completed until shortly before cytoplasmic division (Figs. 28-37). Almost immediately after the bands begin to separate and unwind, the achromatic figure begins to form between the twro groups of separating bands (Figs. 28, 29). In most dividing organisms the fibres of the achromatic figure, like the flagella, appear to arise directly from the flagellar bands about midway between the nuclei and the anterior ends of the cells, but in a few individuals — especially those distorted by drastic treatment — a fine, darkly staining line may be seen following one band of each group (of two bands). This line is the elongate centriole from which the fibres of the achromatic figure arise (Fig. 31). It should be noted here that in other genera of the Spirotrichonymphida1 previously studied (Cleveland, Hall, Sanders, and Collier, 1934) two elongate centrioles follow the spiral flagellar bands from their point of origin at the anterior ends of the cells for a considerable distance posteriorly, but their posterior or distal ends, which produce the achromatic figure, are free of the bands. In Pseudotrichonympha, a hypermastigote without flagellar bands (Cleve- land, 1935), the unusually long centrioles adhere to the inner margin of the rostral portion of the flagellated area, although this, compara- tively speaking, is a short portion of the centrioles; in Barbulanympha, a hypermastigote with two, short flagellated areas at the anterior end instead of flagellar bands, the centrioles which are about 40 microns in length, adhere to the lamella underneath the flagellated areas for a distance of 6 to 10 microns; and in Trichonympha, another hyper- mastigote without flagellar bands, the posteriorly directed elongate centrioles are free except for their point of origin where they are fastened to the lamella underneath the flagellated area. Thus, there are all degrees of anchorage of the centrioles to another organelle, ranging from Trichonympha, Joenia, Joenopsis, and other genera, where the anchorage is slight, through Barbulanympha, Urinympha, Rhynchonympha, Eucomonympha, Teranympha, Staurojoenina, Macro- spironympha, Spirotrichosoma, and Leptospironympha to Spirotricho- nympha, where it is complete. Without this series, the situation in Spirotrichonympha would seem unusual and would be difficult to understand.
LONGITUDINAL AND TRANSVERSE DIVISION
The manner in which the achromatic figure arises from the centri- oles that follow two of the four flagellar bands is similar to that in other hypermastigotes: Astral rays arise from the distal end of each centriole, meet, join, overlap, and grow along one another to form the central spindle (Figs. 28-31). As more rays overlap, the central spindle becomes broader; and as the rays increase in length, it becomes longer. However, a point is reached shortly after nuclear division when the central spindle ceases to broaden, and it soon begins to narrow, because the daughter flagellar-band centriole complexes, to which its ends are anchored, move in opposite directions and thus pull apart the astral rays that formed the central spindle, the last-joined astral rays being the first ones to pull apart (Figs. 33, 35). Finally, the central spindle is pulled apart and disappears (Fig. 36).
The early central spindle lies a considerable distance anterior to the nucleus (Figs. 28-30), but, as it increases in size and the anterior ends of the developing daughter cells move in opposite directions, it is gradually moved toward the nucleus, finally stretching directly across it (Figs. 31, 32). The direction of the central spindle in S. polygyra is crosswise to the long axis of the body, while in S. bispira, as will be described presently, it extends parallel with the long axis. Shortly before the central spindle comes in contact with the ever intact nuclear membrane as a result of its moving posteriorly, some of the astral rays become extranuclear chromosomal fibres by making contact with the intranuclear chromosomal fibres anchored in the nuclear membrane, and as the central spindle increases in length, due to the centrioles from which it arises moving in opposite directions, these extranuclear chromosomal fibres begin to move the chromosomes toward the poles, one short and one long daughter chromosome moving toward each pole (Figs. 9, 10). In order to conserve space and since the same situation exists in S. bispira, where it is illustrated, the extranuclear chromosomal fibres have not been drawn; nor has the relation of the central spindle to the nucleus and chromosomes been shown in most instances. However, the position of the central spindle and of the extranuclear chromosomal fibres can be determined in the drawings of the nuclei by the direction of the intranuclear chromosomal fibres; the extranuclear chromosomal fibres extend from the intranuclear chromosomal fibres in the nuclear membrane in the same direction as the intranuclear chromosomal fibres (Figs. 9-12); and the direction of the extranuclear chromosomal fibres indicates the position of the central spindle, since they and the fibres of the central spindle arise from the distal ends of the centrioles. The nuclei (Figs. 3-15) are all mounted so that their anterior surfaces point toward the top of the page.
10 L. R. CLEVELAND
Just before the chromosomes begin their poleward movement, they become considerably shorter and the turns made by their spirals become broader. These stubby chromosomes persist until shortly after nuclear division (Figs. 9-12); then they gradually elongate and their coils and sheaths split lengthwise, thus producing one short and one long pair of chromosomes which persist through the interphase (Figs. 13-15).
SPIROTRICHONYMPHA BISPIRA SP. NOV.
Morphology
In ten individuals the body ranged in length from 59 to 102 microns with an average of 81; the width at the widest portion ranged from 32 to 48 microns with an average of 40; the distance of the greatest width from the anterior end averaged 57 microns; the distance from the anterior end to the nucleus averaged 21 microns; the transverse diameter of the nucleus averaged 10 microns; the distance from the posterior termination of the spiral flagellar bands to the posterior end of the body averaged 27 microns; the length of that portion of the body covered by the flagellar bands (the flagellated area) averaged 54 microns; the number of turns or spirals made by the flagellar bands averaged 34; the length of the flagella from the surface of the body averaged 20 microns; the length of the flagella from the basal granules to the surface of the body averaged 7 microns; the total length of the flagella averaged 27 microns; the distance from the flagellar bands (straight line) to the surface of the body averaged 5 microns, and the width of a flagellar band is about 1 micron.
It is clear from these measurements that S. bispira differs from S. polygyra neither in the size of the body nor in the size of the organ- elles. And the interphase organelles of the two organisms do not differ in appearance and number except for the number of bands and the enclosure of each pair of daughter chromosomal coils in a common sheath in S. bispira, differences already noted. There is, however, a notable difference in the number of turns made by the flagellar bands, those of S. polygyra making 45 and those of S. bispira making 34. Here is a fact which in itself indicates that the two-banded and four- banded forms are distinct organisms; for, if the forms with two bands making 34 turns developed two more bands and these new bands followed the old ones, as they would have to do, the number of turns .\ "iild IK- f)S. The bands of .S'. bispira are slightly farther apart, >ecially in the anterior end, than those of S. polygyra (Figs. 1,2). The description given for the axostyle, flagella, parabasals, and other • AtnimK -lear organelles of S. polygyra takes care of these organelles
LONGITUDINAL AND TRANSVERSE DIVISION 1 1
in S. bispira, and we may proceed immediately with the description of mitosis.
Mitosis
The various processes connected in one way or another with mitosis and cell division in S. bispira are unusually interesting because they are so plain, because of their unique features, and because they differ decidedly from those of S. polygyra. There is one short and one long chromosomal sheath in the interphase nucleus and each sheath contains two chromosomal coils (Fig. 16). The division of the coils occurs in the late telophase (Fig. 27). Each coil is attached to the nuclear membrane by a fibre which is termed the intranuclear chromosomal fibre. Until the early anaphase, each coil is thus anchored to the anterior margin of the nuclear membrane (Figs. 16—20). (In all the drawings of nuclei the anterior margins are directed toward the top of the page.) The opposite or posterior ends of the chromosomes are all attached either to a single chromatin nucleolus or to portions of a chromatin nucleolus, as described in 5. polygyra.
The first step in nuclear division is the division of the chromosomal sheath and the moving apart of the daughter chromosomal coils (Fig. 17). As the daughter chromosomes move apart, they frequently coil around each other (Figs. 17-19). This movement of the chromo- somes is not poleward movement, but merely movement within the nucleus, and the daughters are probably interconnected during this period — although interconnections cannot always be seen— for later when they contract and straighten, just before poleward movement begins, interconnections may be seen plainly (Fig. 20) ; and these strands connecting the chromosomes are not broken until considerable progress in poleward movement has been made (Figs. 21-23). It should be noted here that poleward movement of the chromosomes is not to the right and left, as in 5. polygyra, when the anterior end of the organism and the anterior surface of the nucleus are foremost, but is anterior and posterior (up and down). In other words, the chromosomes of 5. polygyra move at right angles to the long axis of the cell, while those of S. bispira move parallel with the long axis. The reason for this will be clear when the development and function of the achromatic figure is explained.
So far as I know, there is no cell hitherto described where the behavior of the centrioles during the formation of the achromatic figure is the same as in S. bispira. There are two spiral flagellar bands in each interphase organism, but it has been impossible to determine whether a centriole follows one of these bands prior to the beginning
12 L. R. CLEVELAND
of cell division, because during division a centriole usually follows one band so closely that the centriole and the band appear as one structure. In other words, the centriole cannot be distinguished from the band, and its presence during division, with the exception of a few instances, can be demonstrated only by its function. The elongate centriole which follows the band during division may persist as an elongate centriole during the interphase, just as the centrioles of Barbulanympha and many other genera persist; it may degenerate except for its anterior portion in the late telophase, just as the centrioles of Pseudotricho- nymplia do; or there may be one short and one long centriole in the interphase as in Trichonympha. At present, I see no way to determine which possibility is correct. But, irrespective of whether the posterior portion of one centriole which follows a band during division degener- ates or persists during the interphase, the manner in which it functions during cell division may be demonstrated clearly.
In order to facilitate the explanation of the processes concerned with cell division, it seems desirable to name and label some of the structures involved in the processes. The two interphase flagellar bands which persist and do not change position during cell division are designated parent flagellar bands (p.f.b.). In most organisms a centriole follows one of these bands from its point of origin at the anterior end of the cell for a distance of 4 to 5 turns or spirals, and these two structures are designated centriole parent flagellar band (c.p.f.b.). Occasionally the centriole does not follow the band for more than two turns before it separates from it and continues inde- pendently; then it is labeled centriole (c.). The earliest stage in cell division is the growth of a new flagellar band from the point of origin of the parent flagellar bands posteriorly. This new flagellar band has a centriole which, in the early stages of development, follows it for its entire length, and where the (wo follow each other they are desig- nated centriole new flagellar band (c.n.f.b.). This new band soon breaks loose from its point of origin, becomes free, and migrates posteriorly; then the end which at one time was connected to the parent flagellar bands at their point of origin in the anterior end of the cell is designated the anterior end (a.e.); the other end which, as development progresses, elongates beyond the point where the centriole terminates is designated the elongating end (e.e.). As this new band elongates posteriorly from its point of origin, it develops basal granules from which flagella (/.) grow out. Astral rays (a.r.) arise from the distal portion of the centriole following (he parent flagellar band and soon meet those that arise from the same portion of the centriole that follows the new flagellar band; \\hen they meet, they join, overlap,
LONGITUDINAL AND TRANSVERSE DIVISION 13
grow along one another and thus form the central spindle (c.s.}. The centriole that follows one of the parent flagellar bands may be referred to as a stationary centriole, and the one that follows the new or free flagellar band as a free centriole, although it is not free in a sense because it is attached to the band. Similarly, the astral rays arising from the two centrioles may be referred to (Figs. 52, 53) as astral rays from free centriole (a.r.f.c.) and astral rays from stationary centriole (a.r.s.c.}.
The formation of a new flagellar band — first attached at its point of origin, then free and migrating posteriorly while an achromatic figure is formed between it and one of the (stationary) parent flagellar bands — presents an interesting picture in the mechanics of cell division and renewal of extranuclear organelles. This migrating flagellar band, extending in length as it migrates, evidently has considerable difficulty in finding its way to the posterior end of the cell, for it is sometimes entirely outside the parent flagellar bands and sometimes partly outside and partly inside. Its movement is further handicapped because it is joined to one of the parent flagellar bands by the central spindle arising from its centriole and that of the parent flagellar band. But in most instances it eventually makes its way to the posterior end of the cell and, while so doing, picks up and carries with it two chromo- somes, half of the nuclear membrane, and other nuclear materials. However, it sometimes fails to make this journey, and these instances are particularly interesting from the standpoint of the manner in which the achromatic figure is formed and its function in nuclear division. Two such instances are illustrated: in one (Fig. 52) the new band appears to be caught by the parent flagellar bands near its anterior end; in the other (Fig. 53) the new band appears to be lodged to the side and anterior to the parent flagellar bands. In the first instance, it is difficult to distinguish clearly astral rays arising from the centriole following the new band from the flagella that are arising from the new band, but the astral rays that are arising from the distal end of the centriole that follows one of the parent flagellar bands are long and plain and extend between the spirals of the parent bands to the nucleus (Fig. 52). In the second instance, astral rays may be seen clearly arising from the centriole following the new band; they extend inwardly and the flagella, for the most part, outwardly (Fig. 53). The astral rays that arise from the stationary centriole (a.r.s.c.) extend within the spirals to the nucleus. In neither instance is a central spindle formed, because the centrioles, from which the astral rays forming the central spindle arise, are so placed that the astral rays arising from one centriole cannot meet those arising from the
14 L. R. CLEVELAND
other. In both instances the chromosomes have the appearance of telophase chromosomes, yet the pairs of daughter chromosomes have not moved apart because the proper formation and function of the achromatic figure was upset by the position of the free centriole which adheres to the improperly placed new flagellar band.
The fact that the free centriole in this organism assumes almost every conceivable position during its posterior migration is responsible for the great variation in the appearance of the achromatic figure. For example, the central spindle of the achromatic figure may extend from the free centriole to the stationary centriole entirely within the parent flagellar bands (Figs. 49-51, 55); it may be entirely outside the parent bands (Fig. 47); one portion of it may be inside and the other outside (Figs. 43, 48) ; or, as is more often the case, it may be partly outside and partly inside (Figs. 42, 44, 45). The end of the central spindle adjacent to the stationary centriole is fairly constant in shape when the centriole follows the parent flagellar band closely, because it assumes the shape of that portion of the centriole-band complex from which the fibres arise, and hence is cylindrical (Figs. 50, 51); but when this centriole does not follow the band closely, as occasionally happens, there is nothing to direct the fibres into a cylindrical position and they merely extend posteriorly from the distal portion of the centriole (Figs. 41, 42, 45). The shape of the other end of the central spindle depends entirely on the position of the new flagellar band and the centriole following it. If the centriole and band lie in almost a straight line, the adjacent end of the central spindle is flat or nearly so (Figs. 47, 51); if they lie somewhat in a cylindrical position, the adjacent end of the central spindle is somewhat cylindrical (Figs. 44, 50) ; if they are zigzag or wavy, so is the adjacent end of the central spindle (Figs. 43, 45, 46, 49). The position of the free band and centriole may also be such that only a few of the astral rays arising from the stationary centriole join those arising from the free centriole to form a central spindle outside the parent flagellar bands, while the other astral rays arising from the stationary centriole extend between the parent bands to the nucleus (Fig. 46). Many variations in the numbers of fibres that join or fail to join occur.
The r61e of those astral rays which, by becoming extranuclear chromosomal fibres, function in the movement of the chromosomes to the poles is plain. These rays are longer than those which in other hypermastigotes perform the same function. This is because the nuc-leus lies farther away from the centrioles, the point of origin of the rays, than in most cells. In other words, the astral rays become long bctnre they reach the nucleus. In some instances these extranuclear
LONGITUDINAL AND TRANSVERSE DIVISION 15
chromosomal fibres may be followed all the way from their point of contact with the intranuclear chromosomal fibres in the nuclear membrane to the poles (Figs. 45, 55); in others they are lost among the fibres of the central spindle soon after they leave the nuclear membrane (Fig. 51). There can be no question regarding the fact that they carry one short and one long chromosome anteriorly and one short and one long one posteriorly, because, when the centrioles from which they arise are so situated in the cell that these fibres cannot reach the chromosomes, the chromosomes do not move apart and the nucleus does not divide.
The chromosomes are anchored to the anterior surface of the nuclear membrane until the anaphase, and if there were no inter- chromosomal connections between the daughters of the two pairs, two of the chromosomes would not move at all until the nuclear membrane began to elongate prior to pulling in two. The situation would be very simple, as is occasionally the case when the connections between the daughter chromosomes are broken in the prophase or metaphase: two chromosomes (a long one and a short one) move to the posterior surface of the nuclear membrane, the nuclear membrane elongates, pulls in two, and daughter nuclei are developed. But the more usual thing is for the connections between the daughter chromo- somes to persist until the two chromosomes that are going to form a part of the posterior daughter nucleus are pulled to the posterior surface of the nuclear membrane (Figs. 21, 22). The connections between the posterior ends of the chromosomes are the last ones to break because the pull on the chromosomes is from their anterior ends. Incidentally, we have here an explanation of what produces the so-called equatorial plate stage in chromosomes; the pull on the daughter chromosomes from opposite directions (antero-posteriorly in this organism) causes them to take up a more or less central position within the nucleus. Variations of this stage, of which there are many in different organisms, result from the irregularity in the behavior of the connections between daughter chromosomes, those in some organisms breaking sooner than in others.
Cell Division
The type of cell division in 5. bispira is so unusual that it deserves special attention. Shortly after the division of the nucleus and the disappearance of the achromatic figure, the new flagellar band, which now lies in the posterior end of the cell (Fig. 56), begins to arrange itself in the form of a spiral, the first portion to so arrange itself being the anterior end (Fig. 57). This is the end which in the early stages
16 L. R. CLEVELAND
of mitosis lies adjacent to the anterior ends of the two parent flagellar bands (Figs. 38, 39), the end which does not elongate and which is labeled a.e. in the illustrations of the formation and development of the achromatic figure (Figs. 39-55). It now becomes the anterior end of the daughter cell which is developing at the posterior end of the parent cell (Figs. 56-59). This new band soon becomes arranged in spirals from the anterior to the posterior end (Fig. 58). Meanwhile, both the posterior and anterior developing daughter cells begin to form new axostyles (Figs. 56-58). Transverse cytoplasmic division soon occurs, producing two independent daughter cells; one, the anterior daughter, obtaining a nucleus and all the parent extranuclear organelles intact except the axostyle; the other, the posterior daughter, obtaining a nucleus, the new flagellar band, the centriole adhering to the new band, a small amount of the parent cytoplasm, and the extranuclear organelles such as flagella, parabasals, axostyles, etc. developed before cytoplasmic division. In other words, all the extranuclear organelles of the parent cell except the axostyle, which is resorbed, are carried over into the anterior daughter, and the posterior daughter gets the new organelles. The axostyle surrounds the nucleus in such a manner that the nucleus probably could not divide if this organelle remained intact. So far as I know, the extranuclear organ- dies of flagellates are either resorbed and each daughter develops a new set of organelles or the parent organelles are distributed among each daughter. S. bispira shows a radical departure from either procedure, and one wonders if this is not the beginning of a new stage of evolution in flagellates.
After cytoplasmic division, the posterior daughter, which is considerably smaller than the anterior one, continues the development of its extranuclear organelles. The axostyle increases in length and breadth and the flagellar band extends posteriori},', forming spirals as it does so, although the spirals for some time are not perfectly arranged (Fig. 59). They soon become arranged more or less perfectly (Fig. 60), and presently another flagellar band begins to develop from the point of origin of the existing or first new band. This hand is referred to as the second new band. It extends posteriorly and soon takes its place along the first new band, so that the posterior daughter cell now has two spiral flagellar bands (Fig. 61). These bands increase in length and finally become as long as the parent bands were prior to the beginning of the processes of cell division. As the bands increase in length, they form more flagella and parabasals, the axostyle becomes grown, and the posterior daughter is now indistinguishable from the parent cell from which it originated (Fig. 2). Meanwhile, the axostyle
LONGITUDINAL AND TRANSVERSE DIVISION 17
of the anterior daughter completes its growth, and it, too, becomes indistinguishable from the parent cell from which it originated.
REFERENCES CITED
CLEVELAND, L. R., 1935. The centrioles of Pseudotrichonympha and their role in mitosis. Biol. Bull., 69: 46.
CLEVELAND, L. R., S. R. HALL, E. P. SANDERS, AND J. COLLIER, 1934. The wood- feeding roach Cryptocercus, its protozoa, and the symbiosis between pro- tozoa and roach. Mem. Acad. Arts and Sci., 17: 185.
CUPP, E. E., 1930. Spirotrichonympha polygyra sp. nov. from Neotermes simplici- cornis Banks. Univ. Calif. Publ. Zool., 33: 351.
18 L. R. CLEVELAND
EXPLANATION OF PLATES
The drawings were made with the aid of a camera lucida from material fixed either in Schaudinn's or Flemming's fluids and stained with haematoxylin.
PLATE 1
FIG. 1. Spirotrichonympha polygyra. Entire organism. Note four spiral flagellar bands. The flagella leave the bands in groups of two, and only two groups are shown from each band. The two adhering flagella extend posteriorly from their point of origin almost to the next band; then they turn and continue somewhat posteriorly to the surface of the body where they separate. The spherical parabasals follow the bands from a point just posterior to the nucleus to their termination. Other bodies, smaller than the parabasals, follow the bands from midway between the anterior end and nucleus to their termination. The axostyle extends from the anterior end around the nucleus to the posterior end of the body, terminating in a fine point slightly beyond the body. X 1600.
PLATE 2
FIG. 2. Spirotrichonympha bispira sp. nov. Entire organism. The extra- nuclear organelles differ from those of S. polygyra only in that there are two instead of four spiral flagellar bands. X 1600.
PLATE 3
Spirotrichonympha polygyra
The nuclei are mounted so that their anterior surfaces are directed toward the top of the plate.
FIG. 3. Interphase nucleus. There are four chromosomes. Each chromosome has a single coil lying within a sheath and is anchored to the anterior surface of the nuclear membrane by an intranuclear chromosomal fibre. The posterior end of each chromosome is attached to a portion of the chromatin nucleolus. There are two long and two short daughter chromosomes. The daughters of each group are some- what coiled around each other. X 3800.
FIG. 4. Interphase nucleus. Both pairs of daughter chromosomes are attached to the chromatin nucleolus at their posterior ends; anteriorly they are attached to the nuclear membrane as explained in Fig. 3. X 3800.
1- IG. 5. Interphase nucleus. Here the pairs of daughter chromosomes are con- siderably coiled around each other, the long pair lies over the short pair, and the posterior end of each chromosome is attached to a portion of the chromatin nucleolus. X 2400.
FIG. 6. Interphase nucleus more heavily stained than the three previous nuclei, so that other material in the nucleus besides the chromosomes stains and thus obscures slightly the arrangement of the chromosomes and their connections with portions of the chromatin nucleolus. X 3800.
FIG. 7. Interphase nucleus so heavily stained that chromosomal details cannot be seen. X 2400.
FIG. 8. Prophase nucleus. The daughter chromosomes arc no longer coiled around each other. The coils within each chromosomal sheath are plain and their direction may be followed clearly. X 3800.
FIG. 9. Rather heavily stained nucleus. Each pair of chromosomes is short and stubby, the turns made by the coils within the sheaths are broader and closer together, and the daughter chromosomes are beginning to move apart. The chromo- somal coils here and until the late telophase are not so distinct as in the interphase and prophase. The attachments between chromosomes and port ions of the chromatin nucleolus will be lost presently, not to be renewed until the very late telophase. X 3800.
LONGITUDINAL AND TRANSVERSE DIVISION 19
FIG. 10. The greatly shortened chromosomes have moved apart and a large one and a small one are preparing to line up on opposite sides of the nucleus. X 3800.
FIG. 11. Nucleus has elongated. One large and one small chromosome have been moved to the left side of the nucleus and the other two chromosomes are in the process of being lined up in the same manner. X 3800.
FIG. 12. The nuclear membrane has pulled in two and each daughter nucleus has a pair of dimorphic chromosomes. X 2400.
FIG. 13. Daughter telophase nucleus. Chromosomes are beginning to elongate and the coils within their sheaths are again distinct. X 2400.
FIG. 14. Daughter telophase nucleus. The chromosomal coils are in the process of duplicating themselves by longitudinal division. X 2400.
FIG. 15. Daughter telophase nucleus. The chromosomal coils have divided and the chromosomal sheaths have almost completed the process of longitudinal division. Connections are being made between chromosomes and portions of the chromatin nucleolus. X 2400.
PLATE 4
Spirotrichonympha bispira
The nuclei are mounted so that their anterior surfaces are directed toward the top of the plate.
FIG. 16. Interphase nucleus. Note a short and a long chromosomal sheath, each containing two chromosomal coils attached to the anterior margin of the nuclear membrane by an intranuclear chromosomal fibre. The posterior end of each sheath is attached to a portion of the chromatin nucleolus. X 3000.
FIG. 17. Prophase. The chromosomal sheaths are dividing longitudinally and each sheath contains a chromosomal coil attached anteriorly to the nuclear membrane and posteriorly to the chromatin nucleolus. The daughter chromosomes are coiled around each other. X 3000.
FIG. 18. Prophase. Daughter chromosomes arranged somewhat differently from preceding stage, the anterior attachments of the long pair having moved pos- teriorly, and each pair is attached posteriorly to a portion of the chromatin nucleolus. X 3000.
FIG. 19. Same as the two preceding stages except that each daughter chromo- some is in the process of becoming attached to a single portion of the chromatin nucleolus. X 3000.
FIG. 20. Chromosomes have become short and stubby preparatory to poleward movement. X 4800.
FIG. 21. Early anaphase. The anterior ends of the long pair of chromosomes have made some progress in their poleward movement, one chromosome moving anteriorly, the other posteriorly. The short pair, as is sometimes the case, has been moved almost to the posterior end of the nucleus, although the chromosomes have not begun to separate. This was brought about by an extranuclear chromosomal fibre from the posterior pole (centriole) becoming attached to the intranuclear chro- mosomal fibre of the chromosome destined to go to the posterior pole. This fibre carried both chromosomes with it because the two chromosomes were connected. X 4800.
FIG. 22. A slightly later stage. The anterior ends of each pair of chromosomes have made considerable progress in their poleward movement, while the posterior ends are still connected and have not moved poleward. The short pair lies to the left and separation has progressed farther than in the pair which lies to the right. Note the evagination of the nuclear membrane at the four points where the intra- nuclear chromosomal fibres connect the chromosomes with the nuclear membrane. The evaginations are produced at these points by the pull of the extranuclear chro- mosomal fibres. X 3000.
20 L. R. CLEVELAND
FIG. 23. The chromosomes have turned completely around, a short one and a long one being directed toward each pole (anteriorly and posteriorly). Practically- all of the connections between the daughter chromosomes have been broken. X 3000.
FIG. 24. The chromosomes have moved farther apart and the nucleus has be- gun to elongate. X 4800.
FIG. 25. The chromosomes have moved still farther apart and have begun to elongate. The nuclear membrane is constricting preparatory to being pulled in two. One intranuclear chromosomal fibre has been duplicated. X 3000.
FIG. 26. A later stage. Nucleus greatly elongated and will soon be pulled in two. Chromosomal coils in anterior pair are in the process of duplicating themselves by longitudinal division. X 3000.
FIG. 27. Posterior daughter telophase nucleus. The intranuclear chromosomal fibres and the chromosomal coils have been duplicated. The chromosomal sheaths, while slightly constricted, have not divided. X 4800.
PLATE 5 Spirotrichonympha polygyra
FIG. 28. Early stage in the separation of the four spiral flagellar bands into two groups. As the bands separate, they unwind. A centriole follows one band of each group from the anterior end to the posterior termination of the achromatic figure. Astral rays have arisen from the distal ends of each centriole; some of these rays have met and formed a portion of the central spindle; others, the posterior ones, are in the process of meeting to form more central spindle. X 1500.
FIG. 29. About the same stage as the previous one, but a different view of the two groups of separating and unwinding bands; anteriorly, the astral rays have met to form a portion of the central spindle; posteriorly, they have not met. X 1500.
FIG. 30. A slightly later stage in the separation of the bands and the formation of the central spindle portion of the achromatic figure. The astral rays have met both anteriorly and posteriorly to form the broad, flat central spindle. Note position of nucleus in this and in the two previous illustrations. X 1500.
FIG. 31. Later stage. Central spindle is longer, nearer the nucleus, bands are farther apart, and the unwinding of the bands has progressed considerably. X 1500.
FIG. 32. Still later stage. Central spindle is longer and lies over the nucleus. The bands which the centrioles follow are separated from the other bands. At the ends of the central spindle astral rays are arising from the centrioles and flagella are arising from the bands, but the flagella have been omitted and the astral rays have been omitted except those that joined to form the central spindle. X 1500.
FIG. 33. A much later stage. Nucleus has divided and daughter nuclei have moved a considerable distance apart; many of the astral rays have pulled apart so that the central spindle is now long and narrow. Note the manner in which the bands are unwinding. X 750.
FIG. 33 A. Detail of the anterior end of Fig. 33. Note the relation of the ends of the central spindle to the bands. X 1500.
PLATE 6
Spirotricliotiympha polygyra
Fu;. 34. Vertical view of a stage slightly later than Fig. 32. Note relation of
central spindle to flagellar bands. The flagella are not drawn full length so as to
•id confusing them with the fibres of the central spindle. The astral rays that do
not join to form the central spindle cannot be differentiated from flagella when the
LONGITUDINAL AND TRANSVERSE DIVISION 21
two arise from the same point on the centrioles and flagellar bands. This is a six- banded individual. X 1500.
FIG. 35. An intermediate stage between Figs. 32 and 33 drawn to show that the ends of the central spindle are not flat but semicircular when the centrioles (and flagellar bands) from which the central spindle arises are not straight (as in Figs. 31, 32). In other words, the end of the central spindle has the same shape as the centriole (and band) from which the astral rays composing the central spindle arise. X1500.
FIG. 36. A late stage in the unwinding of the flagellar bands. Achromatic figure has disappeared and daughter nuclei are far apart. New daughter axostyles are growing out and have extended a short distance beyond the nuclei posteriorly. X 1200.
FIG. 37. Elongate organism about to divide longitudinally. The unwinding of the parent bands is complete and new daughter bands will grow out presently. X 1200.
PLATE 7
Spirotrichonyinpha bispira
The flagella arising from the two parent or old bands and those arising from the new or free band are not drawn as long as they are. They are stopped where they leave the body.
Explanation of labels used in Plates 7-10.
a.e. — Anterior end (of new or free flagellar band). a.r. — Astral rays.
a.r.f.c. — Astral rays from free centriole. a.r.s.c. — Astral rays from stationary centriole.
c. — Centriole.
c.n.f.b. — Centriole of (or following) new flagellar band. c.p.f.b. — Centriole of (or following) parent flagellar band. c.s. — Central spindle. e.e. — Elongating end (of new or free flagellar band).
/. — Flagella. p.f.b. — Parent flagellar band.
FIG. 38. Anterior end of a cell in the earliest stage of division showing the twro parent flagellar bands (p.f.b.), the new flagellar band and its centriole (c.n.f.b.), and flagella (/.), arising from the parent bands and the new band. The free centriole, which follows the new band from its point of origin in the anterior tip of the cell, has not begun to produce astral rays; nor has the stationary centriole which follows one of the parent bands. X 1500.
FIG. 39. A slightly later stage. The elongating end (e.e.) of the new band has begun to extend laterally, while the anterior end (a.e.) still remains at the anterior tip of the cell. X 1500.
FIG. 40. Later stage. The new band has begun to migrate posteriorly, its anterior end (a.e.) lying to the left and its elongating end (e.e.) to the right. The free centriole following the new band has given off astral rays which have joined some of those given off by the stationary centriole following one of the parent bands to form the very small, early central spindle (c.s.). In this instance, the centriole (c.) fol- lowing one of the parent bands does not follow the band for its entire length, the distal portion being free. X 1500.
FIG. 41. The situation here is the same as in Fig. 40 except the new flagellar band is longer, has migrated farther posteriorly, the central spindle (c.s.) is longer and wider, and the stationary centriole is free of its band for a greater distance an- teriorly. X 1500.
L. R. CLEVELAND
FIG. 42. Later stage. The new band, especially the elongating end (e.e.) has migrated farther posteriorly and now lies near the nucleus. The central spindle (c.s.) is longer. The distal portion of the stationary centriole (c.) is free of the band. X 1500.
FIG. 43. A considerably later stage in the development of the achromatic figure and the posterior migration of the new flagellar band. The central spindle (c.s.) is in two portions, one inside the parent flagellar bands and one outside. The portion inside the bands extends over the nucleus. No portion of the stationary centriole is free of the band. The shape of the end of the central spindle arising from the centriole following one of the parent bands is the same as the band and centriole from which it arises, semicircular. The other end of the central spindle is irregular in shape due to the position of the centriole from which it arises. Numerous astral rays (a.r.) extend toward the nucleus from the centriole that follows the new band. X 1500.
FIG. 44. Approximately the same stage as illustrated by Fig. 42, but different in appearance because of the position of the new band and its centriole. The end of the central spindle (c.s.) adjacent to the centriole of the new band is cylindrical or nearly so and no portion of the centriole of the parent flagellar band is free. X 1500.'
FIG. 45. About the same stage as Fig. 44, but the achromatic figure presents a very different appearance because the new flagellar band and its centriole lie near the periphery of the cell and extend antero-posteriorly in almost a straight line, thus giving the posterior end of the central spindle a greatly flattened appearance. Note the length of the functioning portion of the centriole of the new band. X 1500.
FIG. 46. The achromatic figure here presents a strikingly different appearance from that of any of the previous illustrations, because many of the astral rays arising from the stationary centriole have extended posteriorly between the spirals of the parent flagellar bands instead of making their way toward the periphery, as a few of them have done, to join the astral rays arising from the free centriole (the one following the new flagellar band). Note astral rays extending from the free centriole toward the nucleus. When these rays join those arising from the other centriole and extending posteriorly within the spirals, another portion of the central spindle will be formed. This portion, like that illustrated in Fig. 43, will lie mostly within the flagellar bands. X 1500.
PLATE 8
Spirotrichonympha bispira
The flagella arising from the two parents or old bands and those arising from the new or free band are not drawn as long as they are. They are stopped where they leave the body. See explanation of Plate 7 for meaning of labels employed.
FIG. 47. The central spindle (c.s.) lies almost entirely outside the parent flagellar bands. Its posterior portion, which arises from the centriole of the new flagellar band, is broad and flat because the new band extends in almost a straight line, while its anterior portion, which arises from the centriole of one of the parent flagellar bands, is semicircular. X 1500.
FIG. 48. Here there are two portions of the central spindle (c.s.), one inside and one outside the spirals of the flagellar bands. The centriole (c.) following one of the parent flagellar bands is free of the band for a short distance. X 1500.
FIG. 40. The central spindle (r.s.) lies within the spirals of the flagellar bands except for the posterior portion which is posterior to the bands. Astral rays arise from both the stationary and free centriole and extend toward the nucleus. Some of these rays have joined the intranuclear chromosomal fibres in the nuclear membrane and by so doing have become extranuclear chromosomal fibres which are pulling the chromosomes toward the poles, the centriole of the parent flagellar band (c.p.f.b.) and the centriole of the new flagellar band (c.n.f.b.). X 1500.
LONGITUDINAL AND TRANSVERSE DIVISION
PLATE 9
Spirotrichonympha bispira
The flagella arising from the two parent or old bands and those arising from the new or free band are not drawn as long as they are. They are stopped where they leave the body. See explanation of Plate 7 for meaning of labels employed.
FIG. 50. Note the manner in which the astral rays composing the anterior half of the central spindle arise from the distal end of the centriole following one of the parent flagellar bands. The central spindle is almost cylindrical at this end because of the spiraling of the portion of the centriole from which it arises, while the opposite end is almost flat. There are probably astral rays arising from the centriole of the new band and extending toward the nucleus but, since they cannot be distinguished with certainty from the flagella which are numerous in this area, they are not drawn. X 1500.
FIG. 51. This central spindle is almost cylindrical anteriorly, flat posteriorly, and twisted in the mid-region. The astral rays that have become extranuclear chromosomal fibres by attaching themselves to the intranuclear chromosomal fibres in the nuclear membrane, as well as the free astral rays, are intermingled with the fibres of the central spindle and can scarcely be distinguished from them, thus presenting a situation closely resembling that of many metazoan mitoses. X 1500.
FIG. 52. In this cell the new flagellar band and its centriole failed to migrate posteriorly so that the astral rays arising from the free centriole (a.r.f.c.) could not meet and join those arising from the stationary centriole (a.r.s.c.) to form a central spindle. X 1500.
PLATE 10
Spirotrichonympha bispira
The flagella arising from the two parent or old bands and those arising from the new or free band, except in Fig. 54, are not drawn as long as they are. They are stopped where they leave the body. See explanation of Plate 7 for meaning of labels employed.
FIG. 53. Another cell where the free band and its centriole failed to migrate posteriorly and the astral rays arising from the free centriole (a.r.f.c.) could not meet and join those arising from the stationary centriole (a.r.s.c.} to form a central spindle. This would be a telophase if the proper formation of the achromatic figure had not been prevented by the position of the centrioles, for the chromosomes are telophasic in structure. X 1500.
FIG. 54. Many of the astral rays arising from the stationary centriole have failed to meet those arising from the free centriole and only a small central spindle connects the two centrioles. The movement of the chromosomes and the division of the nucleus have been upset. The flagella of the new band are drawn full length, but only one flagellum is shown beyond the point of bifurcation. X 1500.
FIG. 55. Nucleus is being pulled in two and the elongate central spindle is pulling apart. X 1500.
FIG. 56. Nucleus has divided, achromatic figure has disappeared, chromosomes have elongated, new axostyles are developing, and the new or free flagellar band has migrated to the extreme posterior end of the parent cell. X 1500.
PLATE 11
Spirotrichonympha bispira
The flagella arising from the two parent or old bands and those arising from the first and second new bands are not shown as long as they are. They are stopped where they leave the body.
24 L. R. CLEVELAND
FIG. 57. The two parent flagellar bands, as in previous illustrations of cell division of 5. bispira, remain intact. The new flagellar band which, although it has assumed various shapes in earlier stages, has not formed spirals, has now begun to spiral at its anterior end, which lies at the posterior end of the cell. The shape of the posterior end of the cell has changed, too. X 1500.
FIG. 58. Later stage. Cell is more elongate, posterior end is pointed more like the anterior end, and all of the new flagellar band has formed spirals. This stage is just before the transverse division of the cytoplasm to form two daughter cells; one, the anterior daughter, getting the two parent flagellar bands and their associated organelles; the other, the posterior daughter which is smaller, getting the new flagellar band and its associated organelles. Both daughters form new axostyles. X 1500.
FIG. 59. A posterior daughter cell after cytoplasmic division. The spirals of what was the new flagellar band before division of the body are little if any better arranged now. This organism has just separated from the anterior daughter. X 1500.
FIG. 60. A slightly later stage in the post -cytoplasmic development of a pos- terior daughter cell. The spirals of the flagellar band have perfected their arrange- ment. X 1500.
FIG. 61. A considerably later stage. A second new flagellar band has now appeared and is co-extensive with the first new band, the two forming spirals in the same manner as those of the parent cell previous to cell division. When these bands increase in length posteriorly (forming spirals as they do), the axostyle extends pos- teriorly, and the body increases in size, this cell will be like the parent which pro- duced'it. X 1500.
HK)I.(X,ICAL I'.I LLETIX
PLATE 1
Onrolhv <». Harris del.
I.. I!
BIOLOGICAL I'.i LI.KTIX
PLATE 2
Dorothy Ci. Harris drl.
I I! Cli-\cl:in,l
BIOLOGICAL HI LLKT1X
PLATK 8
15
Dorothy G. Harris del.
L. R. Cleveland
BIOLOGICAL BULLETIN
PI, ATK 4
25 26
Dorothy (j. Harris del.
L. K. Cleveland
BIOLOGICAL BULLETIN
PLATE 5
26
30
29
31
33
Dorothy G. Harris del.
I.. H. Cleveland
BIOLOGICAL BULLETIN
PLATE 6
34
35
37
36
Dorothy G. Harris del.
L. R. Cleveland
BIOLOGICAL BULLETIN
PLATE 7
42
-V-c.rxfb
46
Of.
Dorothy G. Harris del.
I.. K. Cleveland
BIOLOGICAL BULLETIN
PLATE 8
— c.p.p.b.
c.s.
49
Dorothy G. Harris del.
L. K. Cleveland
HIOLOGICAL BULLETIN
PLATE 9
C.n.fb.-,
Q.r. — /- -z.
Dorolhy G. Harris del.
L. R. Cleveland
BIOLOGICAL BULLETIN
PLATE 10
Dorothy G. Harris del.
L. R. Cleveland
BULLETIN
PLATE 11
57
Bl
58
59
Dorothy O. Harris del.
I.. K. Cleveland
ORIGIN AND DEVELOPMENT OF THE ACHROMATIC
FIGURE
L. R. CLEVELAND
(From Ike Biological Laboratories, Harvard University, Cambridge, Massachusetts)
The question of the origin and nature of the achromatic figure has been debated for more than fifty years, and its existence has been seriously questioned many times, particularly during recent years. Several reasons have been advanced for regarding it as an artifact, the principal one being the inability to demonstrate clearly its origin, development, and function in living cells.
I have studied the achromatic figure in twenty genera of hyper- mastigote and ten genera of polymastigote flagellates. These organ- isms furnish a diversity of favorable material. In some of them, there is nc question regarding the manner in which the achromatic figure is formed from the centrioles and the reality of the fibres composing it; in others, the development of the achromatic figure is not so plain, and can be determined with certainty only in view of their relation to those organisms where the process is unmistakably clear. There are gradations between the two groups. In several genera, the develop- ment of the achromatic figure has been followed in living cells from its very beginning to its disappearance in the late telophase; and the picture presented is exactly the same as in fixed cells.
In one genus, Barbulanympha, considerable time has been devoted to the origin and development of the achromatic figure when more than two centrioles are concerned in the process, and when only one is concerned ; because the process is so very plain under these conditions that no one can question it. In a previous paper (Cleveland, Hall, Sanders, and Collier, 1934. Mem. Am. Acad. Arts and Sci., 17: 185) Barbulanympha was described, and an account was given of the manner in which the two elongate, interphase centrioles produce daughter or new centrioles like themselves, a process which need not be considered here; but it is desirable, before considering the achro- matic figure when an abnormal number of centrioles are present, to consider it when two are present.
The centrioles vary in length in the four species of Barbulanympha from 15 to 30 microns; in the unstained living cell they are of a dense hyaline nature, and may be differentiated easily from other cellular contents; when fixed in Schaudinn's fluid and stained with Heiden-
41
42 L. R. CLEVELAND
hain's haematoxylin, they stain and destain in about the same manner as chromatin ; they are joined at their anterior ends by a desmose; and their distal ends, which lie 20 to 30 microns apart, are free (Fig. 1).
The distal end of each centriole is surrounded by a spherical cen- trosome, which is 4 to 6 microns in diameter, and which moves with the centriole whenever the latter is moved, either by the natural movement of the cell or by mechanical manipulation. When the nucleus and centrioles are suitably destained for study, the centro- comes retain little or no stain.
In the interphase cell, there are no fibres extending from either sentriole (Fig. 1). In the prophase, astral rays begin to grow out from the distal end of each centriole (Fig. 2). At first, the rays arising from one centriole are a considerable distance from those arising from the other; but, as they increase in length, the two sets of astral rays soon meet (Fig. 3), and, as they meet, the individual rays or fibres join, grow along one another and overlap to form the early central spindle portion of the achromatic figure (Fig. 4). That the astral rays arise from the centriole instead of the centrosome may be shown by observation of living cells; the rays may be traced through the cen- trosome to the centriole. The same may be demonstrated also in fixed cells. And further proof that the centrosome plays no part in the production of the astral rays may be obtained from a study of several other genera with elongate centrioles similar to those of BarbulanympJia but with no centrosomes surrounding their distal ends; in these, the connection of the astral rays to the centrioles may be seen at a glance. Those genera with centrosomes surrounding the distal ends of their centrioles form cylindrical central spindles, while those with no centrosomes form flat ones; and thus the role of the centrosome in directing the astral rays that form the central spindle may be seen.
As the astral rays continue to increase in length, more of those arising from one centriole meet those arising from the other, and the central spindle becomes larger; and, at the same time, the central spindle increases in length, because the centrioles from which it arises move in opposite directions (Fig. 5). Those astral rays which do not participate in the formation of the central spindle, because they do not meet and overlap, also increase in length; at the same time, more astral rays arise from each centriole, so that the individual rays vary greatly in length; and it is not long after the central spindle is formed before some of the astral rays which do not participate in its formation become extranuclear chromosomal fibres by joining the intranuclear chromo- somal fibres, the fibres which anchor the chromosomes to the ever intact nuclear membrane (Fig. 6). Thus, the achromatic figure
ORIGIN OF ACHROMATIC FIGURE 43
consists of astral rays which are joined together to form the central spindle, astral rays which are connected to the chromosomes and are responsible for their movement to the poles, and astral rays which merely radiate in the cytoplasm and, so far as known, perform no function. All of these fibres appear alike structurally, and the only justification for naming them is their functional differences.
When three centrioles are present in the interphase (Fig. 7), the type of achromatic figure produced by them during cell division varies considerably, depending on their position. If the distal ends of two of the centrioles lie fairly close together, as illustrated in Fig. 8, two long central spindles and one short one are produced. In this instance, the centriole which lies somewhat apart functions jointly with each of the other centrioles in the production of the two long central spindles; and the two centrioles which lie fairly close together function jointly in the production of a short central spindle between them. In other words, the astral rays arising from the centriole which lies more or less alone, and which may be termed the remote centriole, meet, join, and overlap those arising from the two centrioles which lie fairly close together, and thus the remote centriole functions as much again as each of the other centrioles in the formation of the two long central spindles, because it supplies as many astral rays for central spindle production as both of the other centrioles combined. If, however, as is frequently seen, two centrioles lie adjacent and one apart, the three centrioles function in the production of one central spindle, which has the same appearance as when only two centrioles are functioning in the production of a central spindle (for example, as in Fig. 5). The achromatic figure appears exactly as that of Fig. 8 would appear if the two centrioles on the right were pulled together so that one centrosome touched the other.
If the distal ends of the three centrioles are more or less equidistant, i.e., at the apices of an equilateral triangle, each centriole functions in approximately the same manner in the production of the achromatic figure; three central spindles, more or less equal in size, are produced, and each centriole functions in conjunction with two other centrioles in their formation (Fig. 10). If, on the other hand, the three centrioles are not equidistant, but lie at the apices of a triangle which is wider on one side, only two central spindles are produced — at least at first; a third may be formed later, as the astral rays become longer, if the centrioles between which a central spindle failed to form earlier do not lie too far apart (Fig. 9). Three is the maximum number of central spindles that can be produced from three centrioles.
Many cells with four centrioles have been studied. The four inter-
44 L. R. CLEVELAND
phase centrioles may all he mature, i.e., fully grown, or some of them may be in the process of development, as shown in Fig. 11 ; but they do not function until growth is completed; and when they function, the type of achromatic figure produced by them depends on their position. If they lie side by side in groups of two, one central spindle is produced, although all four centrioles participate in its formation; if they lie at the four corners of a square, six central spindles are pro- duced; if they lie at the corners of a rectangle, four central spindles are usually produced; if they lie fairly close together, but not adjacent, in groups of two, two central spindles are usually produced in the early stages of cell division, and in the later stages, if the centrioles of each group move apart, a central spindle is formed between the centrioles of each group. Sometimes, the four centrioles lie in the same plane in groups of two in such a manner that a criss-crossing occurs in the astral rays forming the central spindles. Figure 13 illustrates an example of this. Here astral rays arising from the upper centriole on the right have joined astral rays arising from both of the centrioles on the left, so that this centriole is connected by a central spindle to each of the centrioles on the left; and the astral rays arising from the lower centriole on the right have joined astral rays arising from both of the centrioles on the left, so that this centriole is also connected by a central spindle to each of the centrioles on the left. In brief, each of the four centrioles is connected to two other centrioles by a central spindle. Would such a series of connections occur as a result of "lines of force"? Other instances have been seen where, for example, the lower centriole on the right was connected to each centriole on the left by a central spindle, while the upper centriole on the right was connected only to the upper centriole on the left by a central spindle. And other vari- ations have been seen in the criss-cross central spindle interconnections of centrioles.
When five non-adjacent centrioles are present, eight central spindles could be formed by an overlapping of the astral rays arising from them, although six, as shown in Fig. 12, is the usual number. Here each of the two upper centrioles is functioning in the formation of three central spindles; the lower right and left centrioles are each functioning in the formation of two central spindles; and the lower middle centriole is functioning in the formation of four central spindles, most of its astral rays being used to form central spindles.
Many examples of a still larger number of centrioles functioning in the production of multiple achromatic figures have been studied, but only one has been illustrated; this is Fig. 16 where fifteen centrioles are in order to avoid confusion only the centrosomes are drawn).
ORIGIN OF ACHROMATIC FIGURE 45
Here there is great variation in the length and width of the central spindles, resulting from differences in the time of their formation and in the number of fibres composing them. There are also many in- stances of the criss-crossing of the fibres of the central spindles; instead of the astral rays arising from one centriole joining those from another, they join those from several centrioles. Few of the central spindles are straight, that is, extend directly from one centriole to another; and it is difficult, owing to their curvation and the manner in which they are crowded together, to determine the number present.
When two centrioles are present but are so widely separated that the astral rays arising from one cannot meet those arising from the other, the central spindle portion of the achromatic figure is not formed (Fig. 14). Similarly, when only one centriole is present, the development of the achromatic figure is incomplete; many astral rays extend to and around the nucleus, but they do not form a central spindle (Fig. 15).
In a few instances where several interphase centrioles are present, decidedly abnormal conditions have been seen; the formation of cen- trosomes surrounding the distal ends of the centrioles is either com- pletely or partially upset, and strands, varying in size from a little larger than astral rays to half the size of the centrioles, extend pos- teriorly from the centrioles (Fig. 17). Just what these strands are is not known. They can scarcely be regarded as astral rays, or even a premature aberrant attempt to produce an achromatic figure; but they may be secondary extensions from the centrioles.
I am indebted to the Penrose Fund of the American Philosophical Society for financial assistance.
BIOLOGICAL BULLETIN
PLATE 1
2
ORIGIN OF ACHROMATIC FIGURE 47
EXPLANATION OF PLATES
All figures were drawn with the aid of a camera lucida from material fixed in Schaudinn's fluid and stained with Heidenhain's haematoxylin. The organism is Barbulanymplia.
PLATE 1
FIG. 1. The two elongate interphase centrioles with centrosomes surrounding their distal ends. X 1400.
FIG. 2. Prophase centrioles with astral rays arising from their distal ends. X 1400.
FIG. 3. The astral rays arising from one centriole are meeting those arising from the other, preparatory to the formation of the central spindle. X 1400.
FIG. 4. The astral rays have met, joined, and grown along one another to form the early central spindle. X 1400.
FIG. 5. More astral rays have joined and the formation of the central spindle is complete. X 1400.
48 L. R. CLEVELAND
PLATE 2
FIG. 6. Longitudinal section of anterior end of cell showing the interrelation of flagellated areas, centrioles, achromatic figure, and chromosomes. Some of the astral rays which do not join and overlap in the formation of the central spindle have become extranuclear chromosomal fibres by connecting with the intranuclear chromo- somal fibres in the nuclear membrane. X 1400.
FIG. 7. Three interphase centrioles and centrosomes. X 1400.
FIG. 8. The type of achromatic figure produced by three centrioles when they lie in this position. X 1400.
BIOLOGICAL BULLETIN
PLATE 2
8
BIOLOGICAL BULLETIN
PLATE 3
10
Dorothy G. Harris del.
1,. K. Cleveland
ORIGIN OF ACHROMATIC FIGURE 51
PLATE 3
FIG. 9. Three centrioles in another position and the achromatic figure produced by them. X 1600.
FIG. 10. Still another position of three centrioles and the achromatic figure produced by them. X 1600.
52 L. R. CLEVELAND
PLATE 4
FIG. 11. Four interphase centrioles, one mature and three immature. X 1400.
FIG. 12. Five centrioles and their achromatic figure. X 1400.
FIG. 13. Four centrioles and their achromatic figure. Note the criss-crossing of the astral rays forming the central spindle. X 1400.
FIG. 14. Two centrioles so far apart that their astral rays could not meet to form the central spindle portion of the achromatic figure. X 600.
BIOLOGICAL BULLETIN
PLAT
I I
\
14
BIOLOGICAL BULLETIN
PLATE 5
15
Dorolhy G. Harris del.
L. K. Cleveland
ORIGIN OF ACHROMATIC FIGURE
PLATE 5
FIG. 15. The type of achromatic figure produced when only one centriole is present. Note the absence of a central spindle. X 1400.
FIG. 16. Fifteen centiioles are functioning in the production of an achromatic figure. In order to avoid confusion, the centrioles are omitted; but the position of their distal ends is indicated by the position of the centrosomes. X 1400.
FIG. 17. Interphase centrioles with probable secondary posterior extensions. X 1400.
THE ADSORPTION OF BACTERIA BY MARINE BOTTOM '
SELMAN A. \VAKSMAN AND UNTO VARTIOVAARA
(From the Woods Hole Oceanographic Institution and New Jersey Agricultural Experiment Station)
In various recent studies (1, 4, 5) on the adsorption of bacteria in soil, the impression is left that the bacteria exert, in that state, only little effect upon the cycle of life in the soil. In his book on Agricul- tural Microbiology, Chudiakov (2) states emphatically that the prob- lem of the condition of the bacteria in the soil hampered considerably a better understanding of soil microbiological processes. The existence of a surface relationship between the soil and microorganisms is true not only of bacteria but also of Protozoa, as shown by Cutler (3).
Different species of bacteria were found to behave differently in regard to their adsorption by the soil particles; their mobility and ability to form zooglea seemed to be of special significance in this connection. The chemical activities of the bacteria, as measured by the evolution of CO2, were considerably modified by the adsorption process (4). The finer soil constituents, namely the clay and silt fractions, were found to have a much greater adsorptive effect than the sand. Different soil types were found to adsorb bacteria to a different degree (5).
Rubentschik (8) made a detailed study of the phenomenon of bacterial adsorption in salt basins; the bacteria isolated from the mud were found to show a higher degree of adsorption than the bacteria found in the water. The conclusion was reached that the bacterial benthos consist of easily adsorbable species whereas the bacterial plankton contains organisms possessing a low degree of adsorption by the bottom sediments. A bacterial exchange took place in the mud, the adsorption of some species being accompanied by the desorption of others from the mud. The relative concentration of the bacteria, their nature and the type of bottom material exerted an influence upon the adsorption process. The activities of different species of bacteria were variously modified in the adsorbed state: the metabolism of some was lowered and that of others was increased.
' ( Ontriliution No. 161 of the Woods Hole OceanoRraphic Institution and J<>urn;il Scries Paper of the New Jersey Agricultural Experiment Station, Department of Soil Microbiology.
56
ADSORPTION OF BACTERIA BY MARINE BOTTOM 57
According to Peele (6), the adsorption of bacteria is probably due to the attraction of unlike electric charges. The nature of the base in the soil complex influences considerably the process of adsorption, the monovalent cations showing the least adsorption.
In most of these studies, pure cultures of bacteria were used and the results obtained interpreted in terms of processes carried out by a complex population of microorganisms inhabiting the soil or the sea bottom. Usually a very short period of contact was allowed between the soil and the bacterial culture grown upon an artificial medium; no attempt was made to determine what happened after the culture had adjusted itself to the environment.
These studies have an important bearing upon the influence of the sea bottom upon the bacterial activities in the water. A number of questions may arise in this connection; 1. Is the specific occurrence of certain bacteria in the bottom due to their adsorption by the bottom material? 2. Is the relatively low number of bacteria in the sea water, as compared with that in the bottom, due to their removal from the water by the adsorption process? 3. Is this relationship responsible for the difference in the rate of bacterial processes taking place in shallow seas over a sand bottom such as Georges Bank, compared with the corresponding processes over a mud bottom, such as the Gulf of Maine (7, 11)?
These and other questions deserve fundamental treatment. No attempt will be made to give a definite answer in the following experi- ments, which must be considered as preliminary in nature.
EXPERIMENTAL
In order to eliminate the interfering effect of the bacterial popula- tion normally found in fresh bottom material, the first experiments were carried out with marine mud and sand which had been kept in a dry state for 5 years and in which the bacteria had been reduced to very low numbers.
In the first experiment, the adsorption of bacteria from a mixed population and from a pure culture of a marine bacterium by two differ- ent types of dry marine mud and sand was studied. A mixed culture was obtained by allowing fresh sea water to remain, in a glass container, in the laboratory, for a period of 48 hours. This resulted in an increase in the numbers of bacteria from a few hundred to 303,000 per 1 cc., as determined by the plate method. As a pure culture, a marine agar liquefying bacterium (No. 11) was used. This organism was selected because of the ease of recognizing the colonies produced on the plate. It had been isolated from sea water and kept in culture for a period
58
SELMAN A. WAKSMAX AXD UXTO VARTIOVAARA
of 2 years. The organism was grown in a medium poor in nutrients (1 gram peptone, 1 gram glucose and 0.5 gram K2HPO4 in one liter of sea water), for 24 or 48 hours. The culture was then diluted ten times with sea water sterilized by heating 30 minutes at 80° C. The two muds, No. 1329 and No. 1331, contained 2.46 and 1.58 per cent organic carbon and 0.28 and 0.16 per cent nitrogen, respectively (9). The sand contained 0.58 per rnit carbon and 0.05 per cent nitrogen. Ten- gram portions of mud or sand were placed in 250-cc. flasks containing either 100 cc. cultured sea water or 100 cc. of the diluted 24-hour culture of the bacterium. The flasks were shaken by hand for 10 minutes, allowed to stand 10 minutes, and 1-cc. portions of the super- natant liquid plated out using a sea water agar medium (No. 1). The
TABLE I
of bacteria from mixed and pure cultures by dry marine mud and sand
|
Nature |
of culture |
l |
Sacteria in 1 cc. |
water, thousand |
s |
|
Cultured u-dti-r Control |
Start 303 |
10 minutes |
2 hours |
21 hours |
|
|
Plus mud |
1329 |
28 |
13 |
' 3,700 |
|
|
Plus mud |
1 531 |
•5 |
75 |
530 |
|
|
Plus sand |
180 |
115 |
32,000 |
||
|
Ii<ii ti-rium No. Control . |
11 |
1,600 |
|||
|
Plus mud |
1329 |
') (,'() |
6 |
5 |
|
|
Plus mud |
L331 |
l.i 000 |
21 |
365 |
|
|
Plus sand |
22.000 |
40,000 |
40,000 |
||
flasks were kept in the laboratory, with occasional shaking, and water plated out again after 2 and 21 hours.
The results presented in Table I show that the pure culture of the bacterium was readily adsorbed by the marine mud. The sand, how- ever, had no adsorptivc effect; there was, in fact, a marked increase in bacterial numbers due to the separation of the bacterial masses or clumps into individual cells as a result of shaking with the sand. In the case of the crude cultures (i.e. the bacterial population in the cultured si -a water), the immediate adsorption of the bacteria was followed by a rapid increase- in numbers. The latter was due to the rapid multiplication of certain bacterial forms at the expense of the organic matter in the mud and sand.
This experiment was repeated and an attempt made to measure not only the multiplication of the bacteria, but also their activities, as iiitlin n< -Ml by the presence of bottom material. This could best be
ADSORPTION OF BACTERIA BY MARINE BOTTOM
59
done by determining the oxygen absorption, as a result of bacterial multiplication (10). Two-gram portions of the two muds and 5-gram portions of the sand were placed in oxygen bottles of about 220 cc. capacity. The bottles were filled with 3-day-old cultured sea water. They were closed and incubated at room temperature, under water and in the dark. At different intervals, some of the bottles were removed and analyzed for bacteria, by plating out 1 cc. of the super- natant water, and for oxygen by the Winkler method.
The results reported in Table II show that there was comparatively little adsorption of bacteria from the mixed bacterial population in cultured sea water, by either marine mud or sand. On the contrary, there was a marked rise in bacterial numbers, reaching a peak in 17 to
TABLE II
Influence of marine bottom material on the growth and metabolic activity of a bacterial
sea water population
|
Nature of culture |
Bacteria in 1 cc. water, thousands |
|||||
|
Sea water control |
Start 0.5 1.8 1.9 2.2 |
3 hours 0.9 1.9 1.7 3.9 |
17 hours 1.2 1,175.0 1,240.0 1,255.0 |
42 hours 1.0 850.0 1,435.0 615.0 |
4 days 7.3 133.0 205.0 65.0 |
S days 3.1 12.2 3.3 3.3 |
|
Plus mud 1329 |
||||||
|
Plus mud 1331 |
||||||
|
Plus sand |
||||||
|
Oxygen consumed, cc. per liter |
||||||
|
Sea water control |
Start 0.0 0.0 0.0 0.0 |
3 hours 0.03 0.25 0.13 0.19 |
17 hours 0.07 0.73 0.62 0.88 |
4% hours 0.23 1.42 1.29 1.57 |
4 days 0.38 2.53 2.47 2.18 |
8 days 0.64 3.48 3.31 3.24 |
|
Plus mud 1329 |
||||||
|
Plus mud 1331 |
||||||
|
Plus sand |
||||||
42 hours, followed by a rapid decline. Active oxygen absorption took place in the bottles accompanying bacterial multiplication; this was due to the oxidation of the organic matter in the bottom material by the bacteria.
A study was now made of the adsorption of bacteria from mixed and pure cultures by the use of relatively large amounts of bottom material. In this experiment, the mud and sand were placed in large test tubes and sterilized by heat at 120° C. for 1 hour. Two volumes of cultured water, 2 days old, or two volumes of a diluted 2-day-old culture of No. 11 were added to the sterile tubes. These were now shaken for 1 minute and allowed to settle for 10 minutes. The super- natant liquid was plated out immediately. The tubes were then allowed to remain in the laboratory for 24 hours; some were shaken
60
SELMAN A. WAKSMAX AND UNTO VARTIOVAARA
occasionally and some were left undisturbed. The water was plated out after 24 hours. The results (Table III) show that there was again a rapid adsorption of the bacteria from the pure culture by the mud but not by the sand. In the mixed culture there was at first a certain amount of adsorption of the bacteria; this was followed by a very rapid increase in numbers. The large amount of mud used offered a good source of energy, especially after heating, for many of the bacteria in the water.
A comparative study was now made of the adsorption of bacteria by fresh marine mud obtained from Buzzards Bay. Varying amounts
TABLE III
Influence of shaking upon the adsorption of mixed and pure cultures of bacteria by
sterile marine bottom material
|
Nature of culture |
Bacteria in 1 cc. of surface water, thousands |
||
|
Shaken |
Unshaken |
||
|
Cultured water control |
4.6 0.3 0.1 KIO.O |
S4 hours 153.0 3,600.0 12,000.0 10,000.0 |
24 hours 170.0 39,000.0 22,000.0 7,000.0 |
|
Plus mud 1329 . |
|||
|
Plus mud 1331 |
|||
|
Plus sand . . |
|||
|
Bacterium No. 11 * control . . . |
33,000.0 3,300.0 30.0 39,000.0 |
214,000.0 1,400.0 1.0 320,000.0 |
130. (1(10.0 31,000.0 1.0 220,000.0 |
|
Plus mud 132<» |
|||
|
Plus mud 1331 |
|||
|
Plus sand .... |
|||
|
Bacterium No. 11 t control |
13,000.0 0.7 30.0 18,000.0 |
21,000.0 1.0 1.0 18,000.0 |
21,000.0 1.0 1.0 25,000.0 |
|
Plus mud 1329 |
|||
|
Plus mud 1331 |
|||
|
Plus sand |
|||
* Dilution of culture with sterile sea water in ratio of 1 t Dilution of culture in ratio of 1 : 30.
10.
of the fresh mud, containing 75 per cent moisture, were placed in 250- cc. Erlcnmcyer flasks. Some flasks received 100 cc. portions of 2 -day- old sea water, while other flasks received 100 cc. of a diluted 2-day-old culture of No. 11. All flasks \vere shaken for 1 minute, allowed to settle for 10 minutes and the supernatant liquid plated. This was repeated after 1, 2, 4 and 10 days. Some of the flasks were kept un- disturbed, while others were shaken at various intervals. The plates were incubated at room temperature for 48 hours and counted.
The results (Table IV) show that fresh marine mud exerted at first .1 reducing effect upon the numbers of bacteria in the mixed population.
ADSORPTION OF BACTERIA BY MARINE BOTTOM
61
After 24 hours, however, there were more bacteria in the water above the mud than in the control water sample. It is interesting to note that the larger the amount of mud used, the higher was the number of bacteria, especially in the shaken mud, in spite of the fact that the immediate reduction in numbers was greater with an increase in the amount of mud. This was due to the fact that the bacteria found in the mud a source of nutrients and began to multiply rapidly. This confirms the results obtained previously by the use of dry mud and sand.
In the case of the pure culture, there was active adsorption of the
TABLE IV
Adsorption of bacteria by fresh marine mud
|
Nature of culture |
Fresh mud * |
Shaking |
Numbers of bacteria in 1 cc. water, thousands |
||||
|
grams per 100 cc. |
Start |
24 hours |
4S hours |
4 days |
10 days |
||
|
Cultured water . |
0 |
0 |
53 |
10 |
9 |
9 |
<1 |
|
Cultured water . |
2 |
0 |
44 |
17 |
4 |
3 |
<1 |
|
Cultured water . |
10 |
0 |
42 |
10 |
32 |
23 |
6 |
|
Cultured water . |
50 |
0 |
24 |
32 |
50 |
13 |
2 |
|
Cultured water . |
0 |
+ |
45 |
16 |
11 |
21 |
<1 |
|
Cultured water . |
2 |
+ |
30 |
13 |
27 |
29 |
17 |
|
Cultured water . |
10 |
+ |
33 |
16 |
23 |
36 |
62 |
|
Cultured water . |
50 |
+ |
23 |
13 |
21 |
79 |
80 |
|
No. 11. |
0 |
o |
12,700 |
140,000 |
240,000 |
236,000 |
310,000 |
|
No. 11 |
2 |
0 |
11,900 |
23,000 |
3,000 |
2,500 |
58 |
|
No. 11 |
10 |
0 |
13,900 |
1,000 |
2,000 |
3,600 |
182 |
|
No. 11 |
50 |
0 |
17,900 |
5,900 |
6,000 |
1,800 |
41 |
|
No. 11 ... |
0 |
+ |
14,700 |
160,000 |
332,000 |
310,000 |
360,000 |
|
No. 11 . |
2 |
+ |
16,600 |
67,000 |
23,000 |
6,700 |
160 |
|
No. 11 |
10 |
1 + |
15,400 |
13,000 |
6,000 |
1,100 |
92 |
|
No. 11 |
50 |
1 + |
15,000 |
4,400 |
1,000 |
68 |
104 |
* Moisture content of mud 75 per cent.
bacteria by the fresh marine mud, but here also the bacteria began to multiply after 24-48 hours. These were not, however, the typical colonies of the agar liquefying organism added to the mud, but various colonies of bacteria commonly found in the marine bottom.
On the basis of these results, one can hardly conclude that the adsorptive action of the mud bottom upon bacteria in the water is detrimental in any manner to their activities. Even though mud adsorbs, removes, or inactivates a large number of bacterial cells such as may be obtained by growing a pure culture of a bacterium upon an artificial medium, there is still no proof that mud as such is in any manner injurious to bacterial processes in general, either those taking
62 SELMAN A. \YAKSMAN AND UNTO VARTIOVAARA
place in the water or in the mud itself. The evidence presented here seems to point merely to the fact that bottom mud exerts a certain controlling effect upon tin- nature of the bacterial population in the sea water. This is, of course, true of the bacteria found in the benthos or in the water immediately above it, but may not hold at all for the plankton bacteria found at great distances from the bottom.
The results of these experiments further emphasize the fact that the phenomena of adsorption obtained by the use of pure cultures of bacteria grown upon artificial media are difficult to interpret in terms of natural marine processes, since the conditions controlling the relations of the mud bottom to the natural mixed population of bacteria in sea water are much more complex.
SUMMARY
1. Marine mud exerted an adsorptive effect upon the bacteria in sea water. This effect was particularly evident when certain pure cultures of bacteria were added to the mud to give a heavy inoculation. Sand bottom had very little adsorptive action upon either mixed or pure cultures of bacteria.
2. In the case of the mixed bacterial population in the water, the adsorption of the bacteria by the mud was soon followed by a rapid increase of bacterial numbers; this took place at the expense of the organic matter in the bottom material.
3. These laboratory experiments showed that certain types of marine bottom material exerted a controlling effect on the numbers of bacteria, which were in intimate contact with them. There was no indication, however, of a permanent paralysis of bacterial growth or metabolism.
REFERENCES
1. CHUDIAKOW, N. N., 1926. t'ber die Adsorption der Bakterien durch den
Hoden und den Einfluss derselben auf die mikrobiologischen Boden- prozesse. Centbl. Bakt., II, 68: 345.
2. CHUDIAKOW, N. N., 1926. Agricultural Microbiology (Russian). Moskau.
Page 214.
3. CUTLER, D. \V., 1919. Observations on soil protozoa. Jour. Agr. Sci., 9: 430.
4. DIANOWA, E. W., AND A. A. WOROSCHILOWA, 1925. The adsorption of bacteria
by soils and its influence upon microbiological activities. Nautchno. Agr. Jour. Moscow, 2: 520.
5. MINENKOW, A. R., 1929. Adsorption von Bakterien durch verschiedene
Bodentypen. Centbl. Bakt., II, 78: 109.
6. PEELE, T. C., 1936. Adsorption of bacteria by soils. Mem. 197, Cornell Univ.
Agr. Exp. Station.
7. KiiusZER, H. W., 1933. Marine bacteria and their role in the cycle of life in
ilu- sea. III. The distribution of bacteria in the ocean waters and muds aboul Cape Cod. Biol. Bull, 65: 80.
8. Ki in NTSCHIK, L., M. B. ROISIN, AND F. M. BIELJANSKV, 1936. Adsorption
of bacteria in salt lakes. Jour. Bact., 32: 11.
ADSORPTION OF BACTERIA BY MARINE BOTTOM 63
9. WAKSMAN, S. A., 1933. On the distribution of organic matter in the sea bottom and the chemical nature and origin of marine humus. Soil Sci., 36: 125.
10. WAKSMAN, S. A., AND M. HOTCHKISS, 1938. On the oxidation of the organic
matter in the sea bottom. To be published.
11. WAKSMAN, S. A., H. W. REUSZER, C. L. CAREY, M. HOTCHKISS, AND C. E.
RENN, 1933. Studies on the biology and chemistry of the Gulf of Maine. III. Bacteriological investigations of the sea water and marine bottoms. Biol. Bull., 64: 183.
PRIMARY SEXUAL PHASES IN THE OVIPAROUS OYSTER
(OSTREA VIRGINICA)
W. R. COE
(From the Osborn Zoological Laboratory, Yale University)
In a recent paper on environment and sex in this species (Coe, 1936a), data were presented to show that the sex ratio at the first breeding season is correlated with age and size at sexual maturity. The proportion of individuals first functioning as females was found to be greatest in those localities and during those seasons most favor- able for rapid growth. The mean size of the young females is larger than that of males of the same age, as Needier (1932a) had previously reported.
During the past year additional evidence has been obtained from other localities along the coast from Long Island Sound to the Gulf of Mexico. This evidence seems to justify certain more or less definite conclusions in regard to the factors which regulate the sexual changes which individuals of this species are known to undergo during their lifetime. Through the courtesy of Dr. Victor L. Loosanoff the writer has had the opportunity of examining additional extensive collections from Milford, Connecticut, while Professor T. C. Nelson has kindly furnished an abundant supply of young oysters of definitely known age from the culture frames of the New Jersey laboratory, on the shore of Delaware Bay, near Cape May. Dr. H. F. Prytherch, of the Bureau of Fisheries, has supplied additional material from Beaufort, North Carolina, and Dr. A. E. Hopkins has sent additional samples from the Gulf Coast laboratory at Apalachicola, Florida. The data concerning the sexual conditions found in this new material are included for com- parison with those from other localities in Table I.
Examination of large numbers of fully adult oysters shows that there is usually an approximately equal number of individuals func- tioning at the time as males and females, or with an excess of females among the older members. The young oysters, on the other hand, always show a large excess of males at their first spawning season. This initial tendency toward maleness is the result of the more rapid growth and multiplication of the male than of the female constituents in the primary reproductive gland in a majority of the young indi- viduals of this alternosexual species (Coe, 1932a, 1934).
From Cape Cod to the Chesapeake Bay, most of the young oysters
64
PRIMARY SEXUAL PHASES IN OVIPAROUS OYSTER
65
may be expected to spawn at the age of one year. North of Cape Cod the first spawning is stated to occur more often at the end of the second year, as is the case with some individuals in certain years south of Cape Cod. On the coast of North Carolina and in the Gulf of Mexico, well nourished individuals of the early set spawn toward the end of their first summer, when only three to four months of age, while those of the later set do not become mature until the following spring.
TABLE I
Observed sex ratios during first breeding season at different localities and in different years
|
Locality |
Date |
No. un- deter- mined |
No. M. |
No. H. |
No. F. |
Total |
Ratio F. to 100 M. |
|
New Haven Harbor |
7-11-32 |
17* |
389 |
4 |
13 |
423 |
3.34 |
|
it U (1 |
7-27-33 |
1* |
129 |
0 |
7 |
137 |
5.42 |
|
it it ( < |
8-2-35 |
105* |
521 |
4 |
40 |
670 |
7.66 |
|
ft ( t It |
7-28-36 |
3* |
327 |
? |
41 |
373 |
12.54 |
|
Milford (Bed) |
8-1-34 |
7* |
80 |
1 |
6 |
89 |
7.50 |
|
it II |
7-20-35 |
?* |
312 |
7 |
7S |
391 |
24.04 |
|
II II |
7-3-36 |
77* |
780 |
10 |
44 |
911 |
5.64 |
|
II II |
8-13-36 |
7 |
410 |
4 |
58 |
479 |
14.15 |
|
II II |
8-25-36 |
I76t |
92 |
0 |
?1 |
289 |
22.83 |
|
Milford (Gulf) |
7-7-37 |
0 |
116 |
3 |
25 |
144 |
21.55 |
|
Milford (Harbor) |
7-7-37 |
10* |
306 |
7 |
57 |
380 |
18.60 |
|
ii ii |
7-28-37 |
6t |
116 |
7 |
79 |
153 |
25.00 |
|
Great South Bay, L. I |
7-31-32 |
21* |
197 |
0 |
7 |
775 |
3.55 |
|
West Sayville (float) . . . |
1932 |
0 |
154 |
4 |
48 |
706 |
31.17 |
|
Delaware Bay |
7-7-37 |
0 |
520 |
13 |
718 |
751 |
41.92 |
|
Beaufort, N. C |
1933 |
0 |
43 |
0 |
71 |
64 |
48.84 |
|
ii ii |
8-10 to 10-20-36 |
10 |
84 |
?, |
34 |
130 |
40.48 |
|
tt ii |
10-20-37 |
73* |
151 |
3 |
56 |
783 |
37.09 |
|
Apalachicola, Fla |
1936 1937 |
81* |
127 |
0 |
Q |
717 |
7.09 |
|
Total |
591 |
4854 |
61 |
809 |
6315 |
16.66 |
* Sexually immature, t Spawned out.
The ratio of the two sexes at their first breeding season varies greatly from year to year and at different localities, as shown in Table I. It also varies at different parts of the season, not only because the males tend to become mature somewhat earlier than the females but also in consequence of an actual change in the sexuality of some of the individuals because a small percentage of the young oysters function as males in the early part of the season and as females immediately thereafter (Coe, 1936a). A few others are hermaphroditic and capable of self-fertilization.
66
W. R. COE
This is well illustrated by three collections of yearling oysters from Milford, Connecticut, kindly supplied by Dr. Victor L. Loosanoff, of the Bureau of Fisheries, in 1936. A sample of 911 yearlings examined at the beginning of the breeding season contained less than 6 recog- nizable females for each hundred males, in addition to 10 hermaphro- dites and 77 small individuals that were sexually immature. At the middle of the season, however, the ratio of females had increased to 14 and near the end of the season to nearly 23 (Table II). In 1937 there was likewise an increase in the proportion of females in the later of two collections from the same locality (Table II). Since the females are often somewhat later in maturing than are the males, they fre- quently spawn later. Hence, if it may be assumed that the immature individuals found early in the season contained a large proportion of potential females and that the majority of those spawned out when the
TABLE II
Change of sex ratio during first breeding season
|
Locality |
Date |
No. undeter- mined |
No. M. |
No. H. |
No. F. |
Total |
Ratio F. to 100 M. |
|
Milford (Bed) |
7-3-36 |
77* |
780 |
10 |
44 |
<M1 |
5 64 |
|
it i* |
8-13-36 |
7* |
410 |
4 |
S8 |
479 |
14 15 |
|
ii ii |
8-25-36 |
176t |
92 |
0 |
2] |
289 |
22.83 |
|
Milford (Harbor) |
7-7-37 |
10* |
Wi |
7 |
S7 |
380 |
18 60 |
|
it l * |
7-28-37 |
6t |
116 |
? |
79 |
153 |
'5 00 |
* Sexually immature, t Spawned out.
last examination was made had previously functioned as males, the differences in the observed sex ratios of the several collections would have an adequate explanation.
An examination of some 4,500 young oysters from Long Island Sound during six successive years has shown a ratio of from 3} to 25 females for each hundred males, depending upon the local conditions and the season (Table I). The average size of the females exceeds that of the males and as a general rule those seasons and situations most favorable for rapid growth have shown the largest percentage of females.
Different culture beds in near-by situations may differ considerably in the ratio of the sexes in the yearling population. Those in shallow, u. inner situations may spawn some weeks in advance of those less laxoialilv situated. The first set may have a more favorable as well
• i liin-er Crowing season than later ones, with a correspondingly
PRIMARY SEXUAL PHASES IN OVIPAROUS OYSTER 67
larger proportion of females. Morphological sex differentiation may occur in the autumn in the one case or it may be delayed until winter in the later sets.
Since there is usually found an approximate equality in the ratio of the sexes or an excess of females after the second year, it is obvious that at least 25 to 45 per cent of the young males in these localities must change to females. It is not yet definitely known whether any of the yearling females change to males but there is conclusive ex- perimental evidence that both males and females sometimes change their sex in the interval between two breeding seasons even at the age of four or more years (Needier, 1932a, Galtsoff, 1937). The stimuli which inaugurate these sexual transformations have not been de- termined, but it is suspected that both inherent tendencies and nutri- tional conditions may be involved.
The cellular mechanism involved in this change of sexual phase has been shown (Coe, 1932a, 1934) to depend upon the presence of gonia of both sexual characteristics or undifferentiated gonia through one sexual phase and their retention as residual cells after spawning. In some species of oysters and other mollusks an alternating rhythm of male and female phases is of regular occurrence (Coe, 1936c).
Summarizing the total collections from each area, as shown in Table I, it may be noted that the collections from New Haven Harbor for the four recent years in which suitable material could be obtained comprised 1,066 males and 101 females, a ratio of 9.6 females for each 100 males. From Milford, only nine miles distant, a total of 2,212 males and 315 females were contained in the eight collections during four summers. The sex ratio would there average 14.2 females for each 100 males, indicating a generally more favorable condition for the young oysters than in New Haven Harbor. This conclusion is sup- ported by the fact that the average size of individuals of the same age was somewhat larger at Milford than at the latter locality.
From a natural bed in Great South Bay, Long Island, an excep- tionally low ratio of females was found in 1932, while from culture floats not far distant, the ratio was nearly nine times as great (Table I) corresponding with a similar increase in the rate of growth.
Farther to the southward the ratio of females is still greater under favorable conditions, averaging more than forty for each one hundred males both on the culture frames in Delaware Bay and at Beaufort, North Carolina (Table I).
68
W. R. COE
CORRELATION BETWEEN SEXUALITY AND SIZE AT FIRST BREEDING
SEASON
The more recent studies are quite in harmony with the evidence presented by Needier (1932a), Burkenroad (1931a), Coe (1932c), Roughley (1933) and others that the mean size of individuals func- tioning as females at the first breeding season is considerably greater than that of males of the same age. In an unselected sample consist-
|
NUMBER OF INDIVIDUAL PERCENT FEMALE 100 90 80 70 60 50 40 30 20 10 0 SIZE mm: |
TOTAL ,S 7 9 32 48 64 110 123 189 81 58 20 10 751 |
||||||||||||
|
/ |
|||||||||||||
|
/ |
|||||||||||||
|
/ |
|||||||||||||
|
/ |
|||||||||||||
|
I |
\ |
/ |
|||||||||||
|
/ |
|||||||||||||
|
/ |
I |
||||||||||||
|
^ |
> |
:• •' |
s |
MEAN' |
|||||||||
|
/ |
|||||||||||||
|
/ |
|||||||||||||
|
15- 20- 25- 30- 35- 40- 45- 50- 55- 60- 65- 70- 19 24 29 34 39 44 49 54 59 64 69 74 |
FIG. 1. Graph showing correlation between sexuality and size at first breeding season, based on an unselected sample of 751 yearlings from the culture frames at the \t-w Jersey laboratory on the Shore of Delaware Bay near Cape May. The size of the class is shown at the bottom of the graph and the number of individuals in each class at the top. The mean percentage of females in the entire sample is also indicated.
ing of 751 yearlings from the culture frames at the laboratory on the shore of Delaware Bay, 358 had a shell length of 50 mm. or more on July 7, 1937, while 393 were smaller. Examination of the gonads showed a ratio of 63.4 females for each 100 males in the group of larger individuals as compared with 26.6 females for each 100 males the smaller ones. This correlation between sexuality and size tin >c 751 young oysters is shown graphically in Fig. 1. It will be noted lltal none of the females was less than 25 nun. in length and
PRIMARY SEXUAL PHASES IN OVIPAROUS OYSTER 69
none of the males more than 69 mm. The largest number of indi- viduals, however, were between 40 and 55 mm. in length and in these groups there was little difference in the percentage of females.
DIRECT FEMALE DEVELOPMENT
The primary gonad may be either distinctly bisexual in appearance or it may become differentiated very early into either of the two sexual phases (Coe, 1932a). In the latter case there is a direct transforma- tion of the primary undifferentiated gonia into the functional sexual cells. The female phase is then attained without indication of protandry.
In Long Island Sound the sexual phase of the first breeding season is sometimes distinguishable as early as October or November in ex- ceptionally well nourished young at the age of only three or four months after setting. More frequently, however, morphological sex differentiation is delayed until mid-winter or later.
At Beaufort, North Carolina, direct female development appears to be the general rule, with a much higher ratio of young females than in more northern localities (Table I). Samples of well nourished individuals from Delaware Bay, known to have set July 9, 1936, like- wise indicated direct development, when examined at the end of December. No evidence of incipient protandry was found at that time, although in many cases the gonads were sexually differentiated.
INDIRECT FEMALE DEVELOPMENT
It has been shown previously (Coe, 1932&), that the primary gonads of some young individuals develop directly into ovaries or spermaries, as the case may be, while those of other individuals of the same age are distinctly bisexual, or, to use a less ambiguous term, ambisexual.
The latter are characterized by a cortical layer of ovogonia and young ovocytes, with spermatogenic cells at one side or in the lumen. In many cases these ovocytes later degenerate or they may remain inactive during the primary functional male phase, to become activated and functional toward the end of that phase or some time thereafter. In this way the sexual phase may change from male to female either during the first breeding season or, much more frequently, during the following autumn.
A subsequent change from female to male phase evidently results from the later activation of some of the descendents of the primary undifferentiated gonia into spermatogonia. The propagation of such undifferentiated gonia may be continued year after year or even through a long lifetime, since some residual gonia always remain after spawning.
70 w. R. COE
It was also previously reported (Coe, 1932a) that the developing gonads exhibit all intergradations between those of the so-called true males in which no indications of ovogonia or ovocytes can be detected and those which develop directly into ovaries. In a few individuals both types of cells in the ambisexual gonad multiply harmoniously, leading to the functionally hermaphroditic condition at the breeding season .
In northern localities some of the spermatogonia in the young ambisexual type of gonad may undergo transformations during the autumn comparable with those taking place in normal spermatogenesis (Coe, 1932a). These cells may later degenerate, followed by the activation of ovogonia, thus leading to the functional female phase at the first breeding season during the following summer. Such incipient protandry seems to be much less frequent than the direct development in which the initial dominant phase, either male or female, is retained during the winter and the following breeding season.
SEXUAL CONDITIONS IN SOUTHERN LOCALITIES
From North Carolina southward the initial sexual phase in the earliest set of the year becomes functional toward the end of the same season, when the young oyster is only three or four months of age, as Burkenroad (193 la) reported for the Louisiana oyster. Young of the later sets are similar to those of more northern localities in their sexual conditions, although protandry may be less frequent (Table I). Thus at Beaufort, North Carolina, and at Apalachicola, Florida, on the Gulf of Mexico, there may be two generations of this species in one year, individuals of the early set spawning some three to four months after setting while those of later broods do not become ripe until the follow- ing spring.
Cultures in both these localities are usually contaminated with more or less numerous individuals of the larviparous species, O. equestris, which matures at a still younger age and has a breeding season covering most of the year in warm situations. It is difficult to distinguish the two species externally when young, but the relations of the epibranchial chamber differ and the gonads are very different, since 0. equestris has a sequence of overlapping male and female sexual phases and the spermatogenic cells are larger than in O. virginica and arranged in dense clusters.
A small collection of the over-wintered young at Beaufort in 1933, kindly supplied by Dr. H. F. Prytherch, showed a ratio of nearly 49 ti males to 100 males, while the first set of the season in 1936, stated to have occurred about June 1, showed a female ratio of about 40 when
PRIMARY SEXUAL PHASES TN OVIPAROUS OYSTER 71
examined in August, September and October. The largest individuals had a shell length of 23 mm. in August, 40 mm. in September, and 70 mm. in October. In August there were only a few males with ripe sperm and toward the end of October most individuals of the early set were spawned out.
Collections of the early set taken September 20, 1937, about fifteen weeks after setting, were from 20 to 50 mm. in length and showed a female ratio of 37.09 (Table I). Those which exceeded 35 mm. in length had a female ratio of about 61 as compared with 16 for those of smaller size.
An unselected sample containing 88 individuals which were one year old, originating from the autumnal set during the last two weeks in September, 1936, on the other hand, showed a ratio of 75 females to 100 functional males. These were doubtless in their second spawning season, having previously participated in the spring spawning during June, 1937, when seven to eight months of age. The ratio of females was thus about twice as great as was found in the young at the first spawning period. This would indicate that the proportion of young males changing to the female phase for their second spawning period must have been about 22 per cent greater than the number of indi- viduals, if any, which experienced a sex change in the opposite direc- tion during the same time.
Small collections from Apalachicola, Florida, obtained through the courtesy of Mr. R. O. Smith, were examined three to five months after setting in 1936. These, combined with an additional collection kindly supplied by Dr. A. E. Hopkins from the same locality in 1937, showed a female ratio of 7.09 (Table I), corresponding closely to the ratios from northern localities under unfavorable conditions.
DISCUSSION
The foregoing evidence concerning the sexual conditions in the Virginia oyster indicates that the sex-differentiating mechanism must be in a very labile condition since it responds so generally to environ- mental influences.
Because of the different sex ratios under different environmental conditions, it seems probable that the responses noted may depend upon the interaction of several associated factors, both genotypic and phenotypic. Among these it is evident that all these populations have at least the three following categories of hereditary and environ- mental influences.
(1). There is an inherent tendency toward protandry, as is the case with so many other mollusks, both pelecypods and gastropods,
72 w. R. COE
including the larviparous oysters (Orton, 1927; Coe, 19326), Teredo (Coe, 1936<f), Venus (Loosanoff, 1937), Patella (Orton, 1928), and Crepidula (Coe, 19366, 1938), to mention only a few of the numerous examples. This tendency may or may not be realized, since direct female development so frequently occurs under favorable conditions.
(2). There appears to be an inherent tendency toward rhythmical change in the sexual phase during successive breeding periods, as in the larviparous oysters, Teredo, and certain gastropods. The fully adult Japanese oyster (Ostrea gigas) was proved by Amemiya (1929) to be subject to this change in sexuality in the interval between two breeding seasons. In the Virginia oyster a corresponding condition has been found by Needier (1932a, 19326), by Burkenroad (1937) and by Galtsoff (1937). The proportion of the adult population in which this change occurred varied greatly in the different experiments, the discrepancies being due presumably both to varying environmental conditions and to comparatively small samples. The most extensive of these experiments is that by Galtsoff, who found that among 202 fully adult oysters of which the functional sexuality was reliably determined during the breeding season of 1936, only 167 retained the same sexual phase in 1937. Of the remainder, 14 had died in the meantime, while 9 or 14 per cent of the females had become functional males, and 11, or 8.9 per cent of the males were found to be in the female phase in 1937. The twenty individuals which experienced this change of sexuality during the interval between the two breeding seasons represented only 10.7 per cent of the total number of survivors.
It is evident, then, that sexual stability may predominate over any tendency toward rhythmical change much more definitely than is the case with the larviparous species. There is also considerable evidence which indicates that the sexual rhythm in the latter is associated with a metabolic rhythm (Coe, 1934). But the nutritive conditions are certainly modified by environmental influences. Change of sexual phase in Crepidula, like metamorphosis in other animals, has been shown recently (Coe, 1938) to be subject to environmental control, and this is presumably true to some extent in the oyster.
Immediately after spawning, the residual gonia in both the sexual phases show considerable variation. Some individuals which have spawned as males show residual small ovocytes in addition to ap- parently undiffercntiated gonia. Others show only spermatogonia and undiffcrentiated gonia, while females may retain few or many or no young ovocytes after the cytolysis of the larger ovocytes. In both sexual phases cytolysis may proceed until only unclifferentiated nia remain and the individual has become essentially neutral.
PRIMARY SEXUAL PHASES IN OVIPAROUS OYSTER 73
Evidently relatively few fully adult individuals have gonads that are so evenly balanced that slight changes in the internal environment may activate either type of the residual gonia and thus determine the sexual phase of the following season. In the majority of both sexes, the same sexual phase is retained (Amemiya, 1929; Needier, 1932a; Galtsoff, 1937; Burkenroad, 1937).
(3). Since the primary gonads in the young oyster before morpho- logical sex differentiation show great variability in the proportion of cells characteristic of the two sexual phase, it seems probable that these are to some extent at least dependent upon different heritable modifying factors (Coe, 1932a, 1934, 1936a). Functional males, hermaphrodites and females may then be expected under uniform conditions, the proportion of each depending upon the combination of factors present in each individual. But in the population of young oysters in any locality preceding the time or primary sexual differentia- tion, in the autumn, winter or summer, according to the locality, even a generally uniform environment will be highly variable in so far as each individual is concerned, since there are differences in age, in situation relative to other individuals and the like (Orton, 1936).
The larger average size of the females at the time of sexual differen- tiation and later and the higher ratio of females in those years and in those localities or situations most favorable for rapid growth may indicate either (a) differences in hereditary factors or (b) a direct response sexually to nutritive and other conditions, or (c) both. The evidence to date seems to imply that both these influences are operative.
The sexual conditions in other species of oviparous oysters are similar to those here discussed for the Virginia oyster. In the Indian oyster (0. cucullata} Awati and Rai (1931) found a corresponding tendency toward protandry, but with an excess of females in the adult population, which averaged 41 per cent males, 3 per cent hermaphro- dites and 56 per cent females. Those which harbored the commensal crab Pinnotheres, however, showed 82 per cent males, 7 per cent hermaphrodites and only 10 per cent females, indicating a masculinizing influence on the part of the commensal, either nutritive or otherwise. Roughley (1933) likewise found a very strong tendency toward pro- tandry in the Australian oyster (0. commercialis) , but the adult population consisted of 54 to 88 per cent females, averaging about 2.7 times as many females as males. He considered that nutritive condi- tions were not influential in determining the sexual phase, since populations under poor conditions for growth showed about the same ratio of females as those having better nutritive conditions. The Portuguese oyster (O. angulata) is another species in which the females
74 W. R. COE
are more numerous than the males in the adult population, Orton and Awati (1926) reporting ratios of 60 to 88 per cent females. Indeed, in relatively few species of fully adult pelecypods are the males as numer- ous as the females (Pelseneer, 1935).
Amemiya's studies on change of sex in the Japanese oyster (1929) include later experiments (1935) in which a portion of the gills was removed from adult oysters immediately after spawning. The opera- tion was performed on four groups of oysters, with an equal number of controls. The following year there was a slightly larger ratio of males in each one of the operated groups than in the controls. It was con- cluded that the proportion of individuals assuming the male sexual phase was influenced by the decreased facility of the operated indi- viduals in obtaining food.
All these considerations emphasize the complexity of genotypic and phenotypic factors which may be influential in controlling the sexual phases of the Virginia oyster. The evidence here presented indicates that the populations along the coast from New England to the Gulf of Mexico consist of several local races living under widely different environmental conditions and differing in some respects in their inheritance. In the more northern localities a certain proportion of the young oysters function as females during their first breeding season at the age of about twelve months either by direct female development or indirectly after passing through an abortive or, less frequently, a brief functional male phase. In southern localities two generations, of which the second comprises a large percentage of females, may occur within the same twelve months period.
LITERATURE CITED
AMEMIYA, I., 1929. On the sex change of the Japanese common osyter, Ostrea gigas
Thunberg. Proc. Imp. Acad. Tokyo, 5: 284. AMEMIYA, I., 1935. Effect of gill excision upon the sexual differentiation of the
osyster (Ostrea gigas). Kept. Jap. Ass. Adv. Sci., 10: 1025. AWATI, P. R. AND H. S. RAI, 1931. Ostrea cucullata (the Bombay oyster). Indian
Zool. Mem., No. 3, Lucknow. BURKENROAD, M. D., 1931a. A study of sex in the Louisiana oyster. In Bull.
No. 1, Dept. Conservation Louisiana, p. 23. BURKENROAD, M. D., 19316. Sex in the Louisiana oyster, Ostrea virginica. Science,
74: 71. BURKENROAD, M. D., 1937. The sex-ratio in alternational hermaphrodites, with
especial reference to the determination of sexual phase in oviparous oysters.
Jour. Marine Research, 1: 75. COE, W. R., 1932a. Sexual phases in the American oyster (Ostrea virginica).
Biol. Bull. 63: 419. COE, W. R., 19326. Development of the gonads and the sequence of the sexual
phases in the California oyster (Ostrea lurida). Bull. Scripps Inst. Oceanogr.,
Tech. Ser.,3: 110. Coi , \\ . l\., l(>32c. Mistological basis of sex changes in the American oyster (Ostrea
virginica). Science, 76: 175.
PRIMARY SEXUAL PHASES IN OVIPAROUS OYSTER 75
COE, W. R., 1934. Alternation of sexuality in oysters. Am. Nat., 68: 236.
COE, W. R., 1936a. Environment and sex in the oviparous oyster, Ostrea virginica.
Biol. Bull., 71: 353.
COE, W. R., 19366. Sexual phases in Crepidula. Jour. Rxper. ZooL, 72: 455. COE, W. R., 1936c. Sex ratios and sex changes in mollusks. Mem. Mus. Hist. Nat.
Belgique (2 sen), fasc. 3: 69. COE, W. R., 1936J. Sequence of functional sexual phases in Teredo. Biol. Bull. 71:
122. COE, W. R , 1938. Conditions influencing change of sex in mollusks of the genus
Crepidula. Jour. Rxper. ZooL, 77: 401. GALSTOFF, P. S., 1937. Observations and experiments on sex change in the adult
American oyster, "Ostrea virginica." Collecting Net, 12: 187. LOOSANOFF, VICTOR L., 1936. Sexual phases in the quohog. Science, 83: 287. LOOSANOFF, VICTOR L., 1937. Development of the primary gonad and sexual
phases in Venus mercenaria Linnaeus. Biol. Bull., 72: 389. NEEDLER, ALFREDA B., 1932a. American Atlantic oysters change their sex. Prog.
Kept. Atlantic Biol. Sta. and Fish. Exp. Sta., 5: 3. NEEDLER, ALFREDA B., 19326. Sex reversal in Ostrea virginica. Contr. Can. Biol.
and Fish., 7: 285. ORTON, ]. H., 1927. Observations and experiments on sex-change in the European
oyster (O., edulis). Jour. Mar. Biol. Ass'n. 14: 967. ORTON, ]. H., 1928. Observations on Patella vulgata. Part I. Sex-phenomena,
breeding and shell-growth. Jour. Mar. Biol. Ass'n. 15: 851. ORTON, ]. H.f 1936. Observations and experiments on sex-change in the European
oyster (Ostrea edulis). Pt. V, A simultaneous study of spawning in 1927
in two distinct geographical localities. Mem. Mus. Hist. Nat. Belgique
(2 sen) fasc. 3:997. ORTON, ]. H. AND P. R. AWATI, 1926. Modification by habitat in the Portuguese
oyster, Ostrea (Gryphaea) angulata. Jour. Mar. Biol. Assn., 14: 227. PELSENEER, PAUL, 1935. Essai d'ethologie d'apres 1'etude des Mollusques. Pub.
Fond. Agathon De Potter. Acad. Roy. Belgique Cl. Sci., Bruxelles, 1-662. ROUGHLEY, T. C., 1933. The life history of the Australian oyster (Ostrea com-
mercialis). Proc. Linnean Soc. New South Wales, 58: 279.
THE DELAYED OCCURRENCE AND TOTAL OMISSION OF
ENDOMIXIS IN SELECTED LINES OF
PARAMECIUM AURELIA
T. M. SONNEBORN (From tJie Department of Zoology, Johns Hopkins University)
Experimental control of the process of endomixis in Paramecium aurelia has been only partly achieved. Jollos (1916), Sonneborn (19376) and others have developed methods of inducing endomixis, but no method of avoiding it completely is known. Temporary suppres- sion of endomixis was probably attained by Jollos (1916); but the method employed volumes of culture medium so great as to make it impossible to ascertain with reliability whether endomixis was occur- ring or not. The present paper sets forth a method that avoids this difficulty and provides not only some lines of descent with greatly extended interendomictic intervals, but also others that never go into endomixis.
The method is based on the observation (Sonneborn, 1937a) that the interendomictic interval varies greatly in sister lines cultivated under the same conditions. An attempt was therefore made to select the lines with the longer intervals. This was done by discarding lines as they went into endomixis and replacing them by new lines begun with surplus individuals from sister lines that had not yet gone into endomixis. When this is done with a group of 12 to 24 daily isolation lines cultivated under the conditions employed by Sonneborn (1936), the following phenomena occur. At first, none of the lines go into endomixis; then, as the "normal" time for endomixis approaches, more and more of the lines go into endomixis and have to be replaced by their sister lines that have not yet gone into endomixis. During a period of several weeks, this high frequency of endomixis continues and necessitates equally frequent eliminations and replacements; but dur- ing the following few weeks the frequency of endomixis greatly de- creases and thereafter occurs but rarely. Selection has thus been effective in seeking out lines in which endomixis completely fails to occur. However, such lines cannot be continued indefinitely; after four or five months they die.
This method has been successfully applied in this laboratory many timo and is now a routine technique. As illustrations of the results <•lii.iin.ible with it, two typical histories will be given in the following
76
AVOIDANCE OF ENDOMIXIS IN PAKAMKCIUM
77
vo -c M
R." ._=>
Q T..T '
£*f"i?
?..? -T..
R
•^ ^ -r
v. d | _
i;
§•
a
.
LI
j-i
ifv. C^ >±. ^ Q 0
t*-, TV-) "^ *O T T
w-t tO >0 O L.I
^ tn V) if) \
J. T T T
<-\j I. . J . .1
_T T. -TN-^ K1 W) -O^ T Q T T* T. -T T-.TT»
o Q a o e, I T |._ 1 1 10*5
£JipiLr M
T..T »
"J T - ,0 ^^M~
UJ T I. -lii —
.J I.' — l.l LJ.ll. I UJ'J>
f'V-Si
tO I .(-, ^o^n ^Dl
— I — '- -
O if, — --- "~|LUUJMJ T
^ - T T TT '•-•
— ul I. I LJ I. ..
UJ -r
*-T-.TT H-l
- .1 LU UJ T-
..J.T.I-
:8l
J.L
u~ CO
^ —
— UJ
1 . .
J T- -' 1 °^
- -.1 0
.y
TT T. J LJ <T O1 CT
^-1 I T ^J •£> '•& LO O
10 I loJ I — — — ••*> o
— I Ivfi I..I Ld UJ LL) ~ ~
^j ro »O uJ |-,l— »O I.-T T- ^" UJ ^
u-, in m T__| uJ m I-J T--T «
— — — I T — le-o S.°S
OD P- ^^ °
Ld C UJ UJ —
-1 =
u,^ ^>!Qs[?- I? 2 I[ oj — — _ IUIQ uj r..J~
— UJUJ *O *LduJ UJ. .2__',. ... 4
|
id |
|
|
j-i ^n ao ° •? 1- -<f •*" *° *M UJ |
Vo ~~" " " ro •? UJ LU 't- LLJ -r I |
LJ
•fS-VSV
•sS ^ <N/T
SlT'TSni
T L..T;3?TJ
J I 1-M-.
|
'1 1 SL |
i |
' j |
t |
u • |
|||||||
o
N
O
,C
CU 4> -w '-
u js e "Tj
3H g g
O C3
CO *L>
u c"0 o>
o «*• *i
r 3
o> • - c £
«J
0)
_c
•§
£ 1
° O &f
-_ a;
Z o
— oj
~U c
-* S CJ -— OJ t« -4-J il J2
cu -r1 •— cj
3 # "° IS
*J U5 0) if
Pll
3 O "" T3
!« -f .Z 'Z
& 0 u ;
^ <U 0) = .=
~ o a c "3 u -c cu c
S S-*
1 3 1 Ji
"~ In J= «
"<5 "^ '> S
u C ? •"
•^ « co ^
9"S o-c ^ = '5
c 15 ,S5 In
1> •« «3 u
.SH
B rt
0)
O co
y 4_t
CJ
'a
o
T) C
O
c
11
&5 ^.2
"o P
t) O
o> -a c c us <u
g s
£3 c
«J -C
o
'Zi J5 o
U O ^2
"? S
aj .-S
K" 5
^ bo
. o
« fe. c 73
^ ****
V^l C/l
3 '-^
o
c ^'
D CU
X t-
UJ 3
O 3 u. CJ
O Of
=: u
v° ° ^ ^
b- •2 § P
u
78
T. M. SONNEBORN
two sections: one of these involves animals of the Johns Hopkins race R, reported upon in recent papers by the author and associates; the other involves animals of Woodruff's long-lived race, here designated W. In the third section, the relation of these results to the variability of the interendomictic interval is set forth.
Race R
Figure 1 gives the results of applying the method of selection to a group of typical lines of race R. The entire group was derived from one individual in endomixis April 14, 1935. The original 12 lines descended from this individual went into endomixis between the 14th and 20th <l,i\> (after 41 to 63 fissions) following the initial endomixis.
TABLE I
The frequency of endomixis and of death in relation to the time since the last preceding endomixis in races R and \V of Paramecium aurelia
|
Days since preceding endomixis |
X umber of lines in endomixis per 100 line-days |
Number of deaths per 100 line-days |
|
|
1-13 |
0 |
0.8 |
|
|
14-55 |
23.3 |
2.6 |
|
|
R |
56-76 |
6.4 |
2.4 |
|
77-122 |
0.6 |
5.9 |
|
|
123-130 |
0 |
35.8 |
|
|
1-30 |
0 |
2.0 |
|
|
W |
31-76 |
3.8 |
0.8 |
|
77-130 |
0.5 |
4.8 |
|
|
131-163 |
0 |
13.4 |
Each of these line's was replaced at the time of endomixis by a new line started with a surplus individual from one of the sister lines that had not yet gone into endomixis. By repeating this process of re- placement in all the descendant lim ^ \\liuirver endomixis occurred or might have occurred, some of the lines lived for as much as 130 days (350 fissions) without endomixis.
As shown in Table I, the life of this group of lines fell into five periods differing markedly in the frequency ot endomixis and death. The first period, extending to the 13th day, was characterized by the complete absence of endomixis. In the second period, from the 14th to 55th days, many lines went into endomixis and had to be replaced. The third period, from the 56th to 76th day, was marked by a great reduction of the number of lines going into endomixis. In the fourth period, from the 77th to 122nd day, only three lines went into cndo-
AVOIDANCE OF ENDOMIXIS IN PARAMECIUM 79
mixis, but the number that died or stopped multiplying increased. Finally, in the fifth period, from the 123rd to 130th day, no line went into endomixis, but the death rate was so high that the group com- pletely died out. In Table I, lines which stopped multiplying and were found, on staining, not to be in endomixis, are included among those which died. Experience showed that this was their usual fate. The main feature of the preceding account is that endomixis did not occur in certain lines of descent carried through 350 fissions during 130 days. This was demonstrated by cytological studies. Each line of descent was stained on every day that fission occurred; and if fission failed to occur in any line for two successive days the line was stained and so brought to an end. Since every line that showed any nuclear condition even remotely suggesting endomixis was discarded, it is certain that no endomixis was overlooked and that the lines living through till the end of the experiment had not experienced endomixis during the entire time.
RACE W
The method of obtaining non-endomictic lines of descent was also applied successfully to race W, though here the detailed results differ from those obtained with race R. A typical set of results on race W is illustrated by Fig. 2. The group there represented began with four individuals in endomixis on February 10, 1935. As shown in Table I, the life of this group is divisible into four distinct periods. In the first period, extending to the 30th day, there was no endomixis and few deaths occurred; in the second period, from the 31st to the 76th day, the frequency of endomixis was at its highest and the death rate was still low; in the third period, from the 77th to 130th day, there were few endomixes (two certainly, and possibly four), but the death rate showed a marked increase; in the fourth period, from the 131st to 163rd day, there were no further endomixes and the death rate reached its peak, resulting in the extermination of the group.
As in race R, the lines of race W which were carried to the end of the experiment showed no endomixis during the entire period of observation. Their life without endomixis extended for 163 days and included 302 successive fissions. This long period is 5| times as long as those reported to be typical for this race by Woodruff and Erdmann (1914) and it includes about seven times as many fissions.
Comparison of the data for the two races given in Table I shows that the group of race W lines had a longer initial period without en- domixis (30 days as compared with 13), and that it at no time at- tained so high a rate of endomixis or death as did the group of race R lines. Further, the maximum period without endomixis was longer in race W than in race R, but included fewer fissions.
|
§ J= 3 •- |
|||||||||||||||||||||||||||||||
|
o |
fvl Q |
» -5 | |
|||||||||||||||||||||||||||||
|
ir |
t |
»o |
" r* ^j |
||||||||||||||||||||||||||||
|
o |
a |
•4-* 5* *^ |
|||||||||||||||||||||||||||||
|
S |
S ^ ^ |
||||||||||||||||||||||||||||||
|
2§ |
P4 |
O |
"2 U £' ££ |
||||||||||||||||||||||||||||
|
T K |
C |
i |
C |
i |
|||||||||||||||||||||||||||
|
> |
| |
> |
tf |
K |
I |
*~_> (*s |
|||||||||||||||||||||||||
|
1 |
• |
j K- |
t |
j |
w o |
||||||||||||||||||||||||||
|
"O |
CM C 0 |
J |
1 |
__ |
'o ^ "« *J c |
||||||||||||||||||||||||||
|
r** |
*o |
<\J |
•S S fc 'So |
||||||||||||||||||||||||||||
|
T |
w |
> |
|||||||||||||||||||||||||||||
|
C p |
G r> - 4 ' |
: ^ |
u h |
1 |
f 1 |
; |
t |
r-^ at |
c — o _O £ -5 |
||||||||||||||||||||||
|
r |
^ to c |
; |
^ T |
1 |
J |
ir> t^J to |
|||||||||||||||||||||||||
|
V |
: |
! |
o |
1 |
O «O CT> |
•— — •— |
|||||||||||||||||||||||||
|
0 ^ c u rO «/» |
rx. (N, -J <NJ Q O m. ) <NJ (\J to O *" 1 Oi 'NJ I «*- o '-- ro *• to |
5 U r L |
| h j |
' |
*( |
J ) tft u uJ u Mf> to f «T ' ..I |
u u r c 7 tC J CNJ i tn L |
•> > J i * u L (, |
8 I-. |
.. |
•| c |
:"=f$5 ) T <r> •-\.::: |
- P • |
* •' o Is ...Jo in T -IV 'J- - M 1 ••• J IT) <"g IT 1 <n |
o ' "Z ^ "o t? 5 t/3 U« l«s o -S-s ^ -w — • <y S ^ a- " o OJ X .~ VI |
||||||||||||||||
|
4 t |
1 |
||||||||||||||||||||||||||||||
|
OJ JT ...... |
|||||||||||||||||||||||||||||||
|
0 |
-J |
- f 1 |
j • |
||||||||||||||||||||||||||||
|
J > |
1 |
a u (< \f |
> 1 j 1 |
||||||||||||||||||||||||||||
|
• |
^ |
- M |
•\ |
ru in . |
<u ^-" o *~" «^ flj |
||||||||||||||||||||||||||
|
•> |
— |
CVJ |
^ _a» n |
||||||||||||||||||||||||||||
|
3 |
, |
i |
•- |
— " "5 |
|||||||||||||||||||||||||||
|
C\J |
«NJ |
° 5 r-N** QJ |
|||||||||||||||||||||||||||||
|
_ _ |
— . |
• ~— |
t* |
O |
j |
*" |
• |
c -s -e |
|||||||||||||||||||||||
|
m |
._ |
<VJ |
1 |
1 |
CSJ |
o .. — |
|||||||||||||||||||||||||
|
t^ |
•P |
fNJ |
C |
; |
N3 ^ _ |
||||||||||||||||||||||||||
|
__ |
*/ |
•) |
"C • j .5 |
||||||||||||||||||||||||||||
|
rsj |
•O |
o .- |
|||||||||||||||||||||||||||||
|
r\i |
-= E -r |
||||||||||||||||||||||||||||||
|
( |
U |
iy> |
CNJ |
IB. |
|||||||||||||||||||||||||||
|
I |
' |
(^ |
rJ |
UJ |
ill |
||||||||||||||||||||||||||
|
• ~ |
^ |
4) «> C d3 |
|||||||||||||||||||||||||||||
|
in |
-C J> OJ u |
||||||||||||||||||||||||||||||
|
CO ^H .N ^ 3 |
|||||||||||||||||||||||||||||||
|
" ^ "O 3 — |
|||||||||||||||||||||||||||||||
|
r^ J§ ° ? |
|||||||||||||||||||||||||||||||
|
-2 S-5 A |
|||||||||||||||||||||||||||||||
|
r*. |
|||||||||||||||||||||||||||||||
|
to |
|||||||||||||||||||||||||||||||
|
CO |
^— ~ ^j '£* '^f ^ |
||||||||||||||||||||||||||||||
|
r^- |
UJ |
^* *"* C/) |
|||||||||||||||||||||||||||||
|
o |
~~ 1 ( ^* Ii^ ^ |
||||||||||||||||||||||||||||||
|
|' |
r^ in r^ |
= 2 8 °"l |
|||||||||||||||||||||||||||||
|
. |
-. |
"-- |
l L § UJ |
, >•"=: ' uj uj o ;p |
R ^^^» g = U .= |
||||||||||||||||||||||||||
|
L |
CM ± ; |
f 0 |
5 g-g i |
||||||||||||||||||||||||||||
|
UJ L |
u ^ |
U 3 § ^ |
|||||||||||||||||||||||||||||
|
UJ _ . |
, L o c ^c |
||||||||||||||||||||||||||||||
|
S C/3 |
|||||||||||||||||||||||||||||||
|
^ (U . — |
|||||||||||||||||||||||||||||||
|
CM |
.. |
re '£ « "§ |
|||||||||||||||||||||||||||||
|
'!• |
= "-">"« |
||||||||||||||||||||||||||||||
|
|| |
J |
o 'S,"^.^^. |
|||||||||||||||||||||||||||||
|
§ |
C 5 " '> |
||||||||||||||||||||||||||||||
|
00 00 co <& |
. J uj |
"" » "2 *o |
|||||||||||||||||||||||||||||
|
aO ^O uj i rv oo °0 T " |
1 |
o <u 5 c • — • u rt • — « |
|||||||||||||||||||||||||||||
|
CQ UJ L UJ T |
J |
• |
.... |
(_> 3 .. ^^ |
|||||||||||||||||||||||||||
|
T..." |
00 UJ |
r*x |
"§ S "5 ^ c |
||||||||||||||||||||||||||||
|
r |
|||||||||||||||||||||||||||||||
|
J |
1 |
^ l fV>" ^" |
|||||||||||||||||||||||||||||
|
n « •.- o 55 g s tr C "~~" **'1 |
|||||||||||||||||||||||||||||||
|
I |
e |
"^ c o -S |
|||||||||||||||||||||||||||||
|
t |
i 9 |
* p «•* *O |
|||||||||||||||||||||||||||||
|
L |
S 10 |
c y (« « ~ c/l l ***^ t/i flj 1 |
|||||||||||||||||||||||||||||
|
a» •- E a> |
|||||||||||||||||||||||||||||||
|
"ai __ 3 J= |
|||||||||||||||||||||||||||||||
|
.§ « C |
|||||||||||||||||||||||||||||||
|
. w J= 0 |
|||||||||||||||||||||||||||||||
|
ia |
o <» aJ H _£ |
||||||||||||||||||||||||||||||
|
>— ' C ' "e UH £ C C |
|||||||||||||||||||||||||||||||
|
JS .2 o |
|||||||||||||||||||||||||||||||
|
O t/J "^ |
80
AVOIDANCE OK KXDOMIX1S IN PARAMECIUM
Three other races (//, E, and S) have been subjected to the same method of selection with similar results. It seems reasonable to sup- pose that what has been accomplished in these five races could prob- ably be done with most or all races <>t P. aurelia under similar cultural conditions: by selection, lines can be isolated that do without endo- mixis throughout their lives — a period 5 to 7 times as long as the ordi- nary interval between successive endomixes.
Variability of the Interendomictic Interval
In the absence of selection, the extent of the period during \vhich no nuclear reorganization occurs was shown by Sonneborn (1937a) to vary greatly. The preceding sections show that still greater vari- ability appears when the selection technique is applied. This greater variability must be due to the existence of relatively rare lines which, without any special experimental treatment, have extremely long interendomictic intervals or even fail completely to undergo endomixis. Since such lines are very rare, they would ordinarily not be found in work where the investigator is able to observe only a small sample of a race. The method of selection serves simply to seek out these lines and multiply them. Without resorting to any change of cultural conditions, it provides a means of obtaining both lines that long omit endomixis and lines that omit it entirely.
SUMMARY
If daily isolation lines of P. aurelia are regularly discarded as soon as they go into endomixis, and if these are then replaced by sister lines that have not yet gone into endomixis, it is possible to maintain for long periods lines which have not been in endomixis since the start of this procedure. With this method there were obtained lines of race R which omitted endomixis for as long as 130 days and 350 fissions, and lines of race W which omitted endomixis for 163 days and 303 fissions. At the end of these long periods, all lines died.
In the culture of groups of lines selected in this way, there is an initial period during which endomixis does not occur; this is followed by a period in which endomixis occurs in many of the lines (which have to be discarded) ; during the remaining history, endomixis occurs rarely and eventually not at all, but the death rate rises and results in the final extinction of the group. The details as to the duration of these periods and their characteristic death and endomixis rates differ in different races. The method of selection emphasizes the enormous variability of the interendomictic interval. The extremely long intervals are very rare, and the selection method is simply a device for finding these rare lines and multiplying them.
T. M. SONNEBORN
LITERATURE CITED
JOLLOS, V., 1916. Die FortpHanzung der Infusorien und die potentielle Unster-
blichkeit der Einzelligen. Biol. Centralbl., 36: 497. SONNEBORN, T. M., 1936. Factors determining conjugation in Paramecium aurelia.
I. The cyclical factor: The recency of nuclear reorganization. Genetics, 21:
503.
SONNEBOR.N, T. M., 1937a. The extent of the interendomictic interval in Parame- cium aurelia and some factors determining its variability. Jour. Exper.
Zool., 75: 471. SONNEBORN, T. M., 1937/j. Induction of endomixis in Paramecium aurelia. Biol.
Bull., 72: 196. XYooDRi'FF, L. L. AM> R. Ki\i)\i ANN, 1914. A normal periodic reorganization process
without cell fusion in 1'aramecium. Jour. Exper. Zool., 17: 425.
THE DISTRIBUTION OF AMMONIA IN THE WATERS OF THE GULF OF MAINE
ALFRED C. REDFIELD AND ANCEL B. KEYS
(From the Woods Hole Oceanographic Institution l and the Biological Laboratories,
Harvard University]
In the decomposition of organic residues in the sea ammonia is apparently the first simple inorganic compound of nitrogen to be set free. Since it does not commonly occur in large quantities, it is evidently oxidized rapidly to nitrate and nitrite. As an ephemeral intermediate in the cycle of regeneration of nitrate, it is of interest in indicating the places in which the decomposition of nitrogenous ma- terial is taking place.
The present paper presents the results of a survey of the occurrence of ammonia in the waters of the Gulf of Maine based on observations made during cruises of the "Atlantis" in September, 1933 and May and June, 1934. The only previous measurements of ammonia in these waters of which we are aware are those made by Seiwell (1931) in Penobscot and Frenchman's Bays and those recorded by Rakestraw (1936) in a study of the occurrence of nitrite in the sea. Our results are correlated with the studies of nitrite made by Rakestraw of the phosphorus cycle by Redfield, Smith and Ketchum (1937), and with observations on the distribution of plankton, based on collections made during these cruises.
METHODS
Ammonia was determined by the Teorell titration of the vacuum distillate as described by Krogh (1934). The apparatus could be used practically on shipboard, using alcohol lamps burning pure alcohol for heating and a Cenco Hyvac pump as a source of low pressure. Am- monia-free water was prepared ashore. It was found necessary to steam out the apparatus before the start of each series of analyses and to make blank analyses between each pair of unknowns. To be reli- able, all determinations must be made in duplicate, since unforeseen contamination frequently occurred.
The sea water samples were taken in Nansen reversing bottle-, from which they were immediately transferred, with double rinsing, to gfess-stoppered bottles. These were stored in the refrigerator and
1 Contribution No. 159.
83
84
ALFRED C. REDF1ELD AND AXCEL B. KEYS
analyzed as quickly as possible. Experiments recorded in Table 1 showed that sea water stored for about eight hours underwent no appreciable change in ammonia content. With longer periods of storage there is a progressive decrease in the ammonia content, which is not prevented by the addition of 0.1 per cent mercuric chloride (Keys, Christcnsen and Krogh, 1935). The reason for this loss is unknown.
RESULTS
The positions of the stations at which reasonably complete and satisfactory observations on the distribution of ammonia were obtained
TAB Lie I
Stability of ammonia in stored sea water samples *
|
1 analysis |
Second analysis |
Third analysis |
|||
|
Hours after collection |
NHj-N, mg./M' |
Hours after collection |
NHj-N. mg./M' |
Hours alter ' dlcction |
NH3-N. mg./M3 |
|
1 |
40 |
7 |
41 |
290 |
10 |
|
2.5 |
11 |
8.5 |
L9 |
290 |
8 |
|
3.0 |
35 |
7 |
35 |
290 |
15 |
|
2.5 |
28 |
8 |
JO |
290 |
2 |
|
4 |
20 |
— |
— |
90 |
13 |
|
4 |
62 |
17 |
51 |
90 |
45 |
|
6 |
-12 |
— |
— . |
90 |
25 |
|
6 |
44 |
18 |
24 |
— |
— |
|
6 |
57 |
18 |
16 |
90 |
37 |
|
5 |
4') |
17 |
28 |
90 |
23 |
|
4 |
46 |
17 |
32 |
— |
— |
|
5 |
42 |
18 |
30 |
90 |
20 |
|
tl.5 |
36 |
290 |
8 |
||
|
t2 |
25 |
290 |
4 |
||
|
t2.5 |
21 |
290 |
15 |
||
|
ts.s |
31 |
290 |
16 |
* All values are averages from duplicate analyses. t I'gCl? added to sample at time of collection.
in September, 1933 and May, 1934 are shown in Figs. 1 and 2. Ob- servations were also made at six stations in the (inll between June 25 and June 30 ; the results obtained al that lime did not differ from those of a month earlier except that on the \\hole smaller amounts oi am- monia were present. In presenting the results it is convenient to separate the observations made in deep water where the stability of the \\.iter column preserves its heterogeneity from those obtained in -hallow water and in the shallower channels across the banks where a degree of turbulence prevails.
DISTRIBUTION OF AMMONIA IN GULF OF MAINE
85
OBSERVATIONS IN DEEP WATER
Figure 3 represents the concentrations of ammonia at various deep stations in September and in May. The stations are arranged as far as possible in the order of the positions through which the water is thought to circulate on penetrating the Gulf. Stations occupied in May are placed under stations occupied in comparable positions in September.
In September, Station 1781, 45 miles southeast of the Eastern Channel, showed the presence of some 10 mg. N per cubic meter in the upper 60 meters, diminishing to zero at 100 meters, below which level no ammonia was detected down to the depth of 2,000 meters. The ammonia in the water in the Eastern Channel was similar in distribu- tion, there being none observable below 125 meters. Within the
FIGS. 1 and 2. Position of stations at which distribution of ammonia was de- termined in September, 1933 and May, 1934 respectively.
Gulf, on the other hand, the concentration of ammonia was distributed rather uniformly throughout the water column. The concentrations of ammonia occurring at any station increased towards the western basin of the Gulf, where amounts exceeding 50 mg. N per cubic meter were found.
In May the distribution of ammonia was very different. Amounts in excess of 10 mg. N per cubic meter were rarely observed in the upper 20 meters and at depths greater than 100 meters. Between these depths, at 30 to 60 meters, maximal concentrations occurred ranging in value from 10 mg. N per cubic meter in the Eastern Channel to 45 mg. N per cubic meter in the southern part of the western basin. In the following month similar conditions were observed save that the maximal concentrations did not exceed 25 mg. N per cubic meter.
86
ALFRED C. REDFIELD AND ANCEL B. KEYS
z cc
UJ
• :
s
•- (/I 5 C
$
r- 0
- 4->
rt
— fc/. o
cn o
UJ
5
|
z |
* |
"> |
||
|
CO |
•. |
<o |
"? |
|
|
• |
N. |
<0 |
||
|
CD |
O |
^ |
||
|
Z |
% |
|||
|
cc |
||||
|
UJ |
o • |
k> |
||
|
I h- |
.*••* •. •- |
* |
;; |
|
|
cc 0 |
i |
•f |
||
|
z |
§ |
" |
• '
|
a: o |
"> |
||
|
UJ •» |
g: i |
.-• |
|
|
H |
K. $^ |
• |
|
|
2 ° < • |
•i |
||
|
UJ |
) |
/ |
|
|
* . . * ^_-^^M^I |
i |
||
|
o |
|||
|
Z -J?l |
|||
|
a: LLI. uj Z' |
yt |
||
|
1- Z |
to * |
||
|
w «tS |
N. O |
||
|
< I |
**" ^ |
||
|
Ul O ' |
• . •• • |
§ i |
ft*^^; |
|
^ |
|||
|
t- »1 |
|||
|
Z UJ |
"\ |
||
|
UJ Q- |
**• 'R |
||
|
Z 00 |
?> , |
||
|
*«» *o |
|||
|
Z" |
• * % |
^ V. |
|
|
o |
i^ — |
||
|
0 2 8 S 8 S 10 2 11? cj p«i |
8 ^^5 o u> O lO 10 9 |
s;
-•
|
•j |
1- |
|
0 |
"» |
|
«\! |
• |
|
. |
|
|
I . s |
. |
|
5 |
j • |
|
|
"• |
j |
|
|
i |
. |
|
|
o |
O O O |
|
|
in |
o >n o |
|
|
CU N ro |
.
o i -
^_ o O»
4, £ =
-C 4J -S
^.= «
tn o
U AJ I
& «"< j5
fc I S
*-• I— 1 O
y
c cs ^
J!S «
<= c 2
c •- c
o £ 2
3 °^
^ !-5 —
'C x E
-b o c
10 d m
.SUM
t a £ JJ £• a
r-* C X
o <u re
:il
c h c
DISTRIBUTION OF AMMONIA IN GULF OF MAINE
87
OBSERVATIONS IN SHALLOW WATER
Figure 4 shows the distribution of ammonia at shallow stations off Cape Cod and on Georges Bank and in the strong tide ways of the South and North Channels. No definite difference in the seasonal picture is evident. In the channels considerable concentrations of ammonia occurred at all depths, though in May a maximal concentra- tion a short distance below the surface was present in the North
20
100
I50J
NO. 1777 IX- 8- 3 3
BOTTOM
20
40
20
40
60
NO 1802 IX- 14-33
NO 1742 IX- 3- 33
20
40
> . ... • -•
NO I7S9 IX- 11-33
50
100
150
NO 2193 V - 28-34
NO 2165 V- 21 - 34
NO 2214 VI- 2- 34
NO 22OI V- 29 - 34
NORTH CHANNEL
CAPE COD
SOUTH CHANNEL
GEORGES BANK
FIG. 4. Vertical distribution of ammonia in shallow waters of the Gulf of Maine. Upper row, in September, 1933; lower row, comparable positions in May, 1934. Concentrations of ammonia expressed as milligrams ammonia nitrogen per cubic meter along the abscissa.
Channel. In the shallow stations the concentration sometimes in- creased with depth, sometimes the reverse.
DISCUSSION
The distribution of ammonia appeared very irregular when the observations were first made. The magnitude of the concentrations of ammonia observed agreed well with the general range of observa- tions made by earlier workers in various regions (Robinson and Wirth, 1934). It appeared, however, that great differences might be observed
ALFRED C. REDFIELD AND ANCEL B. KEYS
in similar situations at any time and in the same situation at different times.
In several ways our observations support the conclusions drawn from similar studies of ammonia in the English Channel made by Cooper (1933). He considers the formation of ammonia to be a "surface" phenomenon with lesser activity at the bottom. In the spring the surface concentrations decrease from utilization by phyto- plankton. Our May observations would appear to be adequately explained on these assumptions. He also observed a general increase in summer and autumn, as did we in September.
As additional evidence has accumulated concerning the biology and chemistry of the Gulf, a number of correlations appeared which seem to give the findings additional significance.
RELATION TO PHOSPHATE CYCLE
The most striking difference between the occurrence of ammonia in September and in May is its relative abundance in the deep water at the end of summer compared with its absence there in the spring. This is precisely the relation found for another intermediate product of decomposition — dissolved, organic phosphorus compounds. These were found by Redfield, Smith and Ketchum (1937) to disappear from the water of the western basin in mid-winter and gradually to ac- cumulate during the summer, reaching a maximum observed concen- tration in November. Like the ammonia in September, the concentra- tion was quite uniform at all depths. In the spring these phosphorus compounds appear in the subsurface layers, as does the ammonia, before they may be detected in considerable quantity in the depths. It will be of interest to observe whether the ammonia does not almost entirely disappear from the water in mid-winter, as does the dissolved organic phosphorus.
Like the organic phosphorus, ammonia appears to accumulate at all depths in late summer. This should not be thought of as the accumu- lation of an inert product, however, for von Brand, Rakestraw, and Rcnn (1937) have shown, at least in a laboratory experiment, that the ammonia of decomposing plankton is nitrified completely in 30 days. It is probable that the ammonia is formed by decomposition throughout the water column, for in September there are no well- marked gradients of ammonia concentration which are necessary for its transport by eddy conductivity. There is no evidence, except in a few of the shallow stations, of increasing concentrations near the bottom, such as Seiwell (1931) observed in shallow bays and as might if decomposition on the sea bottom was responsible for a con- part of the regeneration of nitrogen.
DISTRIBUTION OF AMMONIA IN GULF OF MAINE
89
RELATION TO NITRATE
If the view to which Rakestraw (1936) has recently lent support is correct — that nitrite arises in the water through the oxidation of ammonia — one would expect a close correlation in the distribution of these nitrogen compounds. This expectation is in part fulfilled. The distribution of ammonia in May closely parallels that of nitrite, as shown in Fig. 5. Both may be explained through decomposition taking place in the subsurface waters. Rakestraw explains the absence of nitrite from the water immediately below the surface as being due to
100
10
20
40
50
150-
200-
AMMONIA
FIG. 5. Vertical distribution of nitrite and ammonia at Station 2212 May 31, 1934. Depths measured in meters downward along the ordinate. Concentrations measured in milligrams nitrogen present as nitrite or as ammonia per cubic meter.
the assimilation of this substance by plankton. Evidently ammonia is similarly consumed. It is interesting, though not surprising, that there is no evidence that soluble organic phosphorus compounds are assimilated in a similar way — since they were found in as high con- centration in the surface as at any depth.
Rakestraw remarks that the maximum for nitrite occurs between June and September. However, the nitrite does not appear to ac- cumulate toward the end of summer to the extent to which ammonia and organic phosphorus compounds do. Nitrite is evidently a shorter-
00
ALFRED C. REDFIELD AND AXCEL B. KEYS
lived link in the chain of nitrogen transformations than is ammonia for even in May the ammonia nitrogen concentration exceeds that of nitrite nitrogen by ten-fold.
HIGHEST
AMMONIA
SEPT 33
HIGHEST AMMONIA
VOLUME OF ZOO-PLANKTON SEPT '34
VOLUME OF
ZOO- PLANKTON
MAY-JUNE '34
FIG. 6. The highest concentration of ammonia in the upper 80 meters of water in various parts of the Gulf of Maine in September, 1933 and May, 1934 compared with the quantitative distribution of zociplankton. Ammonia concentrations ex- pressed in milligrams ammonia nitrogen per cubic meter. Zooplankton volumes expressed as cubic centimeters of dry plankton per square meter of surface area caught in a vertical haul from near the bottom.
RELATION TO PLANKTON
The quantity of ammonia observed during both cruises varied !^i rally from station to station. In September the greater quantities
DISTRIBUTION OF AMMONIA IN GULF OF MAINE 91
were observed in the northern and western parts of the Gulf; in May in the southern and eastern regions. These facts did not assume significance until the plankton catches made at the time had been analyzed.
During each cruise a standard vertical zooplankton haul was made at each station occupied (a number far exceeding those at which the ammonia was investigated). A 1.5-meter silk net was hauled from near the bottom to the surface. The total catch at each station was freed of excess moisture on a filter and its volume measured by dis- placement. The results of these measurements expressed as the vol- ume of "dry" plankton per square meter surface have been plotted in Fig. 6 to show the general distribution of animal plankton. For comparison, the highest value of ammonia observed at each station in the upper 80 meters is also shown. While the correlation is not per- fect, it is evident that the highest concentrations of ammonia at both periods of observation occur in regions in which the zooplankton are densely distributed.
This correlation suggests that the animal plankton may be in some way responsible for the appearance of ammonia. While specific observations on plankton animals are lacking, ammonia appears to be the principal nitrogenous waste product of many invertebrate animals (Delauney, 1931). Harvey (1934) has pointed out that the phyto- plankton community is continuously being grazed down by the zo- oplankton. These considerations point to the probability that zooplankton, through their metabolic products and through their own decay, are an important intermediary in the liberation and distribution of ammonia in sea water.
SUMMARY
1. In the deeper basins of the Gulf of Maine in May ammonia occurred in minimal concentrations at the surface and at all depths below sixty meters; maximal concentrations varying up to 45 mg. N per cubic meter occurred in a definite stratum between 30 and 60 meters.
2. In September the concentration of ammonia was rather uniform at all depths and increased as the distance from the open sea increased, concentrations exceeding 50 mg. N per cubic meter occurring in the western basin.
3. In the tideways of the North and South Channels, ammonia is distributed uniformly with depth in both May and September.
4. In shallow waters its occurrence showed no regularity.
5. The occurrence of ammonia may be correlated in part with the distribution of organic phosphorus compounds, of nitrite, and of
92 ALFRED C. REDFIELD AND ANCEL B. KEYS
zooplankton, so as to support the view that its distribution marks the place and the intensity of organic decomposition.
REFERENCES
VON BRAND, T., N. \V. RAKESTRAW, AND C. E. RENN, 1937. Biol. Bull, 72: 165.
COOPER, L. H. N., 1933. Jour. Mar. Biol. Ass'n., 18: 677.
DELAUNEY, H., 1931. Biol. Rei'.t 6: 265.
HARVEY, H. W., 1934. Jour. Mar. Biol. Ass'n., 19: 775.
KEYS, A., E. H. CHRISTENSEN, AND A. KROGH, 1935. Jour. Mar. Biol. Ass'n.,
20: 181.
KROGH, A., 1934. Biol. Bull., 67: 126. RAKESTRAW, N. VV., 1936. Biol. Bull., 71 : 133.
REDFIELD, A. C., H. P. SMITH AND B. KKTCHUM, 1937. Biol. Bull., 73: 421. ROBINSON, R. J., AND H. E. WIRTH, 1934. Jour. Conseil. perm. int. pour I'explor.
de la mer, 9: 15. SEIWELL, H. R., 1931. Ecology, 12: 485.
THE OXYGEN CONSUMPTION OF ARTIFICIALLY ACTI- VATED AND FERTILIZED CH^TOPTERUS EGGS
JEAN BRACKET
(From the Marine Biological Laboratory, Woods Hole, Mass., Biology Department, Princeton University, and the Faculty of Medicine, University of Brussels)
INTRODUCTION
It is well known that unfertilized Chsetopterus eggs, when they are treated with various physical or chemical agents, undergo a special type of development (F. R. Lillie's (1902, 1906) differentiation without cleavage). Lillie's original observations have been recently confirmed and completed by Pasteels (1934) and by myself (1937). These inves- tigations show that if the unfertilized eggs are treated with 5 per cent isotonic KC1 in sea water, they first complete their maturation; im- mediately after the expulsion of the second polar body, they start undergoing a series of monaster cycles. During each of the cycles, there is an increase in the number of the chromosomes and the appear- ance of lobulations simulating cleavage. But the furrows fade out quickly and the egg resumes its original shape. There is, therefore, a considerable increase in the size of the egg nucleus and in the number of the chromosomes during the first hours of differentiation without cleavage. Later on, the big single nucleus breaks down and, at the same time, occur processes comparable to gastrulation; in the best cases, unicellular ciliated eggs, resembling somewhat trochophores in their general shape, can be obtained.
It seemed of interest to measure the oxygen consumption of eggs undergoing differentiation without cleavage and to compare it with the respiration of unfertilized eggs: it was hoped that such an inves- tigation would throw some light on the question whether energy is needed for cleavage and differentiation, a problem which has recently been investigated by Tyler (1933, 1936) on other eggs.
A beginning was made last year during a stay in the Zoological Laboratory in Naples and a report has been published recently (1937). The results, however, were far from satisfactory because a small per- centage only of the eggs could be fertilized, although differentiation without cleavage was easily obtained. Lacking the necessary controls, we could only compare the data obtained for the activated eggs with the results published by Whitaker on the respiration of fertilized eggs in Woods Hole. Obviously, no definite conclusion can be drawn from
93
94 JEAN BRACKET
such a comparison and it was therefore necessary to reinvestigate the problem.
MATERIAL AND METHODS
The experiments were carried out following the technical details pointed out by \Vhitaker (1933) : one or two ripe females were allowed to shed after the tip of their parapodia had been cut with scissors. The eggs were then washed repeatedly (5 or 6 times) in sea water in order to remove the mucus and a suspension of a concentration varying from 1 : 25 to 1 : 40 was prepared by slightly packing the eggs in a graduated centrifuge-tube and adding the required amount of sea water. Aliquot parts of the suspension (2 cc.) were pipetted into Warburg manometer conical cups (ca. 10 cc. capacity). Fertilization with one drop of diluted sperm or activation with 0.1 cc. of isotonic KC1 usually occurred just before attaching the vessels to the manometers. Readings were taken after a 15-minutes equilibration period. The water-bath was kept at 24.8° and the manometers were shaken at a speed of 50 round- trip oscillations per minute with an excursion of 4 cm. That these conditions may be considered as adequate is indicated by the fact that increase in the speed of shaking induced no increase in the oxygen consumption ; furthermore, the respiration of unfertilized eggs usually kept constant during 6 or 7 hours and they sho\ved at that time little, if any, cytolysis.
RESULTS
We kno\v since Whitaker's experiments that there is a strong drop in the oxygen consumption of Chxtopterus eggs at the time of fer- tilization. My previous experiments from Naples indicated that a similar drop occurs in KCl-activated eggs. In most of those experi- ments, activation was produced by tipping the KC1 contained in the side-arm into the main part of the vessel after the respiration had already been measured during 1 or 2 hours. The results at Naples could be easily confirmed at Woods Hole: while the metabolism of the fertilized eggs dropped to 53 per cent of the initial level (average from 14 experiments in which the percentage of fertilized eggs exceeded 90 per cent), the oxygen uptake of the activated eggs (from the same females) fell to 51 per cent. There is therefore no doubt that, like the increase in respiration which follows fertilization in sea-urchin eggs, the drop observed in the present case is linked to the cytoplasmic changes resulting from activation and not to any special influence exerted by the spermatozoon on the metabolism.
The following graph represents the oxygen consumption of Chsctop- terus eggs during the 7 hours following activation and fertilization.
OXYGEN CONSUMPTION CH^TOPTERUS EGGS 95
In this graph, the rate of oxygen consumption has been plotted against time, the initial respiration of the unfertilized eggs being considered as 100 per cent. The individual points on the curves represent average values obtained from 14 experiments.
Rate of 00 consum-pttoa **•%
200r
-150
-100
50
u,
•i 204507 Hours.
FIG. 1. Oxygen consumption of Chxtoptems eggs during the seven hours fol- lowing activation and fertilization. Ordinates: rate of oxygen consumption; ab- scissae, hours. The curve F represents the oxygen consumption of fertilized eggs; the curve KC1 represents that of eggs treated with KC1; and the circles, u, represent that for unfertilized eggs.
It is easy to see that, as stated before, the oxygen consumption of the unfertilized eggs (indicated by large circles) remains constant throughout the whole experimental period. It is evident too that respiration increases after the drop in oxygen consumption occurring after fertilization or activation, but obviously this increase is much stronger and faster in the case of the fertilized eggs than it is for the activated eggs: the fertilized eggs reach the initial level after 3y£ hours, i.e. at an early blastula stage. The respiration rate increases steadily later on and is almost 80 per cent higher at the end of the experiment than the oxygen uptake of the unfertilized eggs. On the other hand, the activated eggs differentiating without cleavage reach the initial rate only after 6 hours instead of 3>^. The fact that the same dif- ference in the respiration of two sets of eggs was found regularly in all the experiments indicates beyond doubt that the observed fact is really significant.
It was possible that the difference might be linked to a depressive effect on the metabolism of the added KC1 which was present during the whole experiment. In order to check this possibility, 4 series of
96 JEAN BRACKET
control experiments were run in which the respiration of activated eggs was measured both in presence of KC1 in excess as before and after repeated washings (10 times). In other manometers, the oxygen consumption of normal fertilized eggs was compared with the metab- olism of similar eggs to which 5 per cent isotonic KC1 was added 10 minutes after fertilization. All these control experiments showed that the presence or the absence of KC1 in that concentration does not exert any significant effect on the curves. In both cases, the initial respiratory level was reached after 3^2 hours in the case of the fertilized eggs and after 6 hours for the activated ones. It seems, however, possible that KC1 slightly enhances the absolute Q% consumption of the eggs. Obviously, the differences observed in the metabolic rate of eggs differentiating without cleavage and of fertilized eggs is some- how linked to their different type of development and not to a de- pressive effect of KC1 on the metabolism.
It \vas also of some interest to see ho\v the thymonucleic acid synthesis, as a chemical index of the mitotic activity, would compare in fertilized eggs and in eggs differentiating without cleavage. The eggs were therefore taken out of the manometer vessels and preserved in acetone until a sufficient amount of material (ab'out 2 grams) could be collected. The thymonucleic acid content of unfertilized, activated and fertilized eggs was determined by Dische's (1930) colorimetric micromethod. Three estimations could be made. These showed that while the unfertilized eggs contained, as was expected, only traces of thymonucleic acid, the activated eggs had a content in this substance of about 0.35 milligrams per gram wet \veight and the fertilized eggs of 1 milligram per gram wet weight. The thymonucleic acid synthesized in 8 hours during differentiation without cleavage amounts thus only to 30 per cent of the amount produced in the fertilized eggs during that period.
DISCUSSION
The curve we obtained for fertilized eggs closely resembles the one published by \Yhitaker (1933), although its slope is somewhat steeper. In Whitaker's experiments, the unfertilized egg rate is reached by the fertilized eggs in 4^ hours and the increase over that rate at the end of 7 hours of the experiment is only 30 per cent. This difference is very likely due to the higher temperature (24.8° instead of 22°) at which our experiments were carried out: ciliary activity had just begun after 6 hours in XYhitaker's case while we noticed swimming blastulae after 4*/£ hours. It seems therefore probable that conditions similar to those described by Tyler (1936a, b) in his work on temperature coefficients of developmental processes and cellular respiration prevail
OXYGEN CONSUMPTION CH^TOPTERUS EGGS 97
also in Chxtopterus eggs. It is very probable that the reduced meta- bolic activity of the eggs differentiating without cleavage is, likewise, somehow linked with their slower development. \Ye have already seen that the control experiments rule out the possibility that the observed difference should result from an inhibition effect of KC1 on the eggs' respiration. There is also no doubt that the development of the eggs differentiating without cleavage is slowed down to a consider- able extent: for instance, at the end of the experiments the fertilized eggs had turned into actively swimming larvae while the activated eggs develop cilia only much later. Likewise, the increase of nuclear material, as indicated by the thymonucleic acid estimations, goes on at a much slower rate in the eggs differentiating without cleavage than in the fertilized ones. It is therefore likely that both the reduced oxygen uptake and the slower development are linked together; such a conclusion could support Tyler's opinion that part of the energy available in the egg is needed for the growth and differentiation processes taking place during development. Recent findings by Runnstrom (1933), J. Brachet (1935), Privolnev (1936), Stefanelli (1937) that the respiration undergoes cyclic changes during mitosis in eggs of different species are also in good agreement with our observa- tions. It must be pointed out, however, that the interpretation of the results is complicated by a special factor; namely, that the activated eggs remain unicellular while the fertilized ones cleave into many cells. It is, of course, by no means impossible that oxidation processes might occur at the surfaces between the different blastomeres. Against such an interpretation, however, two facts can be cited; namely, that there is no direct relationship between the number of the blastomeres and the oxygen uptake of the egg (cf. Needham, 1931) and that the respira- tion of small marine invertebrate eggs is over a wide range independent of the O2 tension in the surrounding medium.
SUMMARY
1. The oxygen consumption of unfertilized Chaetopterus eggs drops in the same proportion whether they are fertilized or activated with KC1.
2. The increase in respiratory rate after that initial drop is much faster in fertilized eggs than in activated (differentiating without cleavage) eggs.
3. The lower respiratory activity of the activated eggs is linked to their special type of development.
98 JEAN BRACKET
BIBLIOGRAPHY
BRACKET, J., 1935. Etude du mctnbolisme de 1'oeuf de Grenouille (Rana fusca) au
cours du developpement. II. La respiration de 1'oeuf pendant la feconda-
tion et la mitose. Arch, de Biol. 46: 1. BRACKET, J., 1937. La differentiation sans clivage dans 1'oeuf de Chetoptere
envisagce aux points de vue cytologique et metabolique. Arch, de Biol.,
48: 561. DISCHE, Z., 1930. Cber einige neue charakteristische Farbreaktionen der Thymonu-
kleinsaure und eine Mikromethode zur Bestimmung derselben in tierischen
Organen mit Hilfe dieser Reaktionen. Mikrochemie, 8: 4. LILLIE, F. R., 1902. Differentiation without cleavage in the egg of the annelid
Chaetopterus pergamentaceus. Arch.f. entw. Mech,, 14: 477. LILLIE, F. R., 1906. Observations and experiments concerning the elementary
phenomena of embryonic development in Chaetopterus. Jour, exper. Zool.,
3: 153.
NEEDHAM, J., 1931. Chemical Embryology. Cambridge University Press. PASTEELS, J., 1934. Recherches sur la morphogencse et le determinisme des seg- mentations inegales chez les Spiralia. Arch. d'Anat. microsc., 30: 161. PRIVOLNEV, T. I., 1936. Ueber den Atmungsrythmus bei der Furchung von Fluss-
prickeneier (Lampetra fluviatilis). C. R. Ac. Sc., U. R. S. S. 4: 433-436. RUNNSTROM, J., 1933. Stoffwechselvorgange wahrend der ersten Mitose des See-
igeleies. Protoplasma, 20: 1. STEFANELLI, A., 1937. Prime osservazioni sulle assunzione di O^ della uova et dei
premi stadi embrionale dei Bufonidi. Bull. Soc. ital. Biol., 12: 284. TYLER, A., 1933. On the energetics of differentiation. I. A comparison of the
oxygen consumption of "half" and whole embryos of the sea urchin.
Publ. Staz. zool. Napoli, 13: 155. TYLER, A., 1935. On the energetics of differentiation. II. A comparison of the
rates of development of giant and of normal sea-urchin embryos. Biol.
Bull., 68: 451. TYLER, A., 1936a. III. Comparison of the temperature coefficients for cleavage
and later stages in the development of the eggs of some marine animals.
Biol. Bull., 71 : 59. TYLER, A., 19366. IV. Comparison of the rates of oxygen consumption and of
development at different temperatures of eggs of some marine animals.
Biol. Bull,, 71 : 82. \YHITAKER, D. M., 1933. On the rate of oxygen consumption by fertilized and
unfertilized eggs. IV. Chaetopterus and Arbacia punctulata. Jour. Gen.
Physiol., 16: 475.
ON THE ENERGETICS OF DIFFERENTIATION. VII
COMPARISON OF THE RESPIRATORY RATES OF PARTHENOGENETIC AND
FERTILIZED URECHIS EGGS
ALBERT TYLER AND N. H. HOROWITZ
(From the William G. Kerckhoff Laboratories of the Biological Sciences, California Institute of Technology, Pasadena, California)
The results of these experiments show that the rise in respiratory rate that occurs during development is correlated with cleavage; eggs that fail to cleave after activation show a greatly retarded rise while those that cleave show a rise that is roughly commensurate with their division rate. Inhibiting cleavage with phenylurethane affects the respiratory rate similarly.
THEORETICAL PART
It has been long been known that the rate of respiration rises during development. This increase in rate is evidently not directly propor- tional to the increase in the number of cells (cf. Needham, 1931). It might, nevertheless, depend upon changes in the egg brought about by cell division, so that when cleavage fails to occur the rise in respiration would be inhibited. We have considered in previous work the dependence of the form-changes on the respiration, the rate of oxygen consumption being taken as a measure of the energy available for the various developmental processes. We consider now the possibility that the developmental changes determine in turn the rate of respira- tion. If, for example, early cleavage is inhibited in a manner that does not affect the absolute rate of respiration at the particular stage, then we may expect, on this basis, failure of the subsequent rise.
In the early work of Warburg (1910) it has been shown that cleav- age could be suppressed in sea urchin eggs by means of phenylurethane without immediately affecting the respiratory rate. However, the question of whether or not the rate would rise later was not investigated. Also it has been shown that after parthenogenetic activation of sea urchin eggs the same increase in rate occurs that is obtained normally upon fertilization (Warburg, 1910; Loeb and Wasteneys, 1913). But here again it would be desirable to know what happens later, especially since the parthenogenetically activated eggs develop much more slowly in general and often stop in early cleavage or even fail to divide.
99
100 ALBERT TYLER AND N. H. HOROWITZ
In Urechis artificial activation with a single agent may produce dividing or non-dividing eggs depending on the length of treatment. Diluted sea water (Tyler, 1931) or ammoniacal sea water (Hiraiwa and Kawamura, 1936; Tyler and Bauer, 1937) may be used for this
purpose.
EXPERIMENTAL PART
The respiration measurements were made by means of the usual Warburg method using the cylindrical type of vessel previously described (Tyler, 1936) but of 18 to 20 cc. calibration volume. The quantity of eggs present in each vessel was determined at the end of the run from the Kjeldahl nitrogen, and the oxygen consumption is expressed on that basis. The quantity of eggs employed was such as to give readings of 25 to 50 mm. Brodie fluid per hour at the start of the experiment.
The non-dividing and the dividing parthenogenetic eggs were produced simply by treatment with ammoniacal sea water, as pre- viously described. Treatments of 2 to 7 minutes with 0.01 N NH3 in sea water give 100 per cent activation with normal polar body extrusion but no cleavage. Such eggs go through a series of monaster cycles after polar body formation, but at a slower rate than would correspond to the normal nuclear changes. After about 10 to 15 hours unicellular swimmers may develop similar to the embryos differentiating without cleavage, first described by Lillie (1902). Treatments of 12 to 17 minutes with 0.01 N NH3 in sea water give 100 per cent activation with as much as 90 to 100 per cent cleavage. Polar body formation is usually interfered with in such eggs, as previously described (Tyler and Bauer, 1937). After 8 to 12 hours swimming embryos including some normal ones may appear.
Non-Cleaving Parthenogenetic Eggs
The results of two sets of experiments with non-cleaving activated eggs are given in Table I, along with the fertilized controls. In both the treatment was for six minutes with 0.01 N NH3, and insemination of the control was done at the start of the parthenogenetic treatment. Activation and fertilization were 100 per cent in both these experi- ments. As may be seen in Table I the oxygen consumption of the parthenogenetic eggs during the first hour of measurement is very nearly the same as that of the fertilized eggs. In both experiments the parthenogenetic eggs give slightly higher values, but the difference is no greater here than the difference between the duplicate vessels. During the subsequent hours the rate rises steadily in the case of the Irrtili/rd eggs. The parthenogenetic eggs, however, show first a
RESPIRATORY RATE AND CLEAVAGE
101
slight drop (second to fourth hours) followed by an increasing rate. After 8 hours the parthenogenetic eggs attain a rate that is about double their initial respiration, but only about half of that of the fertilized eggs.
The eggs from the vessels were examined at the end of the run. In both experiments the parthenogenetic eggs gave about 2 per cent of unicellular swimmers while the fertilized eggs gave 90 to 98 per cent normal top-swimming trochophores.
Cleaving Parthenogenetic Eggs
The respiration data of two sets of experiments with eggs that divide after ammonia treatment are given in Table II. In both
TABLE I
Oxygen consumption of non-cleaving parthenogenetic eggs. Parthenogenetic treatment = 6 minutes with 0.01 N NH3 in sea water. Measurements begun at 43 minutes after treatment or insemination in Experiment I and 60 minutes after in Experiment II. Values given as cu. mm. C>2 per hr. per mg. N. Temperature = 22.0° C.
|
Experiment |
[ |
II |
|||||
|
Hour |
Parth. |
Parth. |
Fert. |
Pert. |
Parth. |
Pert. |
Pert. |
|
1st |
4.64 |
4.48 |
4.38 |
4.24 |
4.72 |
4.53 |
4.14 |
|
2nd ... . |
4.40 |
4.30 |
4.53 |
4.38 |
3.88 |
5.05 |
4.49 |
|
3rd |
4.02 |
3.99 |
5 60 |
5 55 |
3.83 |
5.45 |
5.35 |
|
4th 5th |
4.52 5.22 |
4.34 5.07 |
7.58 10.01 |
7.53 9.93 |
4.32 5.56 |
7.01 10.10 |
6.65 10.20 |
|
6th |
6.19 |
6.02 |
13.60 |
13.43 |
7.40 |
13.83 |
13.48 |
|
7th |
6.86 |
7.29 |
16.38 |
16.31 |
9.66 |
18.26 |
17.45 |
|
8th |
8.91 |
8.87 |
18.66 |
18.47 |
10.93 |
21.64 |
20.81 |
experiments it may be seen that the respiration of the parthenogenetic eggs at the start is slightly higher than that of the fertilized controls. A third set not presented in the table shows the same initial difference, and also agrees very well throughout the run. The difference here is somewhat greater than was manifested in the experiments with the non-cleaving parthenogenetic eggs. It may be pointed out that the ammonia treatment is considerably longer here (13 to 17 minutes) and this may account for the higher initial rate.
The subsequent readings give values that are fairly constant for the parthenogenetic eggs up to the fifth hour, while the fertilized eggs rise as usual during this time. Following this the rate rises and at the eighth hour the respiration is about half of the control rate. It is very nearly the same at that time as is obtained with the non-cleaving
102
ALBERT TYLER AND N. H. HOROWITZ
parthenogenetic eggs (Table I). However, differences appear later as will be shown below.
The cleavage of these parthenogenetic eggs is very slow compared with the fertilized controls. In Experiment I, at 2 hours after treat- ment, there were 43 per cent cleaved of which 15 per cent were in two cells, 22 in three and 6 in four. The fertilized controls were 100 per cent in four at this time. An hour later there was 75 per cent cleavage. At the end of the run (11 hours after treatment) the eggs from the vessels were 90 per cent cleaved and 30 per cent were swimmers.
In Experiment II, 60 per cent of the parthenogenetic eggs were divided at two hours after treatment, the distribution being 23 per cent
TABLE II
Oxygen consumption of cleaving parthenogenetic eggs. Parthenogenetic treatment
= 13 minutes (Experiment I) and 17 minutes (Experiment II) with 0.01 N NH3 in
sea water. Measurements began at 74 minutes after treatment or insemination in
Experiment I and 53 minutes after in Experiment II. Values given as cu. mm. O*
per hour per mg. N. Temperature = 22° C.
|
Experiment |
] |
[ |
I |
I |
||||
|
Hour |
Parth. |
Parth. |
Pert. |
Pert. |
Parth. |
Parth. |
Pert. |
Unfcrt. |
|
1st |
5.88 |
5.81 |
4.48 |
4.64 |
5.69 |
5.70 |
4.51 |
3.69 |
|
2nd |
5.45 |
5.69 |
4.92 |
4.82 |
5.40 |
5.48 |
4.60 |
3.90 |
|
3rd |
5.20 |
5.25 |
5.63 |
5.80 |
5.51 |
5.34 |
5.59 |
3.37 |
|
A 4-U |
51 £. |
51 A |
7A7 |
A OA |
5QC |
c 70 |
f\ QJ. |
|
|
4tn |
.oO |
.v>0 |
.\)o |
o.yo |
.80 |
o. / U |
||
|
5th |
5.63 |
5.25 |
10.28 |
10.12 |
5.91 |
5.48 |
9.07 |
3.29 |
|
6th |
5.99 |
5.92 |
14.05 |
13.80 |
7.01 |
6.86 |
13.17 |
3.37 |
|
7th |
7.55 |
7.25 |
17.46 |
17.83 |
8.04 |
8.02 |
16.27 |
|
|
8th |
8.62 |
8.04 |
20.90 |
21.14 |
10.25 |
10.18 |
20.06 |
|
|
9th |
9.71 |
9.49 |
24.50 |
24.45 |
||||
|
10th |
10 65 |
1095 |
2855 |
|||||
in two cells, 30 in three and 7 in four. The fertilized controls were all in four cells at the time. At the end of the run (10 hours) the par- thenogenetic eggs from the vessels were 95 per cent cleaved of which 5 per cent were swimmers, while the fertilized eggs gave 90 per cent swimmers.
Since the parthenogenetic eggs do not divide at all synchronously and since many may stop after one or more cleavages, the distribution becomes quite complicated. It is clear, however, that cleavage and the development of cilia are retarded.
Later Stages
Measurements on the later stages were made by culturing the eggs in dishes and washing the embryos before transfer to the vessels.
RESPIRATORY RATE AND CLEAVAGE
103
This is to avoid possible effects of bacterial growth or other changes
produced in the vessels during a prolonged run. In Table III the results of one set of experiments are given. Cleaving and non-cleaving
parthenogenetic eggs were prepared from the same batch along with the fertilized controls, and allowed to develop for 11 hours at room temperature (18.2° C.) before transfer to the vessels. In the fertilized lot there were more than 95 per cent top-swimming young trocho- phores, and only the top swimmers were transferred to the vessels. In the non-cleaving parthenogenetic lot (Parth. 5) there were no swimmers at that time. In the cleaving parthenogenetic lot (Parth. 17) 95 per cent had divided and practically all of them were bottom- swimmers.
TABLE III
Oxygen consumption of late stages of non-cleaving and cleaving parthenogenetic eggs. Parthenogenetic treatment = 5 minutes (Parth. 5) and 17 minutes (Parth. 17) with 0.01 N NH3 in sea water. Eggs cultured for 11 hours at 18.2° before measure- ments were begun. Values given as cu. mm. O2 per hour per mg. N. Temperature = 22°.
|
Hour |
Fert. |
Pert. |
Parth. 5 |
Parth. 5 |
Parth. 17 |
Parth. 17 |
|
1st |
21.80 |
22.09 |
12.31 |
12.17 |
11.64 |
11.69 |
|
2nd |
25.45 |
25.91 |
12.54 |
12.42 |
19.03 |
18.45 |
|
3rd |
29.38 |
30.48 |
15.11 |
14.90 |
18.74 |
18.96 |
|
4th |
32.39 |
33.32 |
15.01 |
14.65 |
20.75 |
20.26 |
|
5th |
36.17 |
36.18 |
14.44 |
14.15 |
22.60 |
22.35 |
|
6th |
39.94 |
40.48 |
15.76 |
15.72 |
24.76 |
24.98 |
|
7th |
43.31 |
45.70 |
18.98 |
18.75 |
27.70 |
27.95 |
|
8th |
46.50 |
47.70 |
22.18 |
22.73 |
31.73 |
32.67 |
As may be seen from Table III the two types of parthenogenetic eggs respire at very nearly the same rate at the start of the experiment. The rate is a little more than half of the fertilized rate at this time. The rate rises with time, but more rapidly in the case of the dividing eggs. At the end of the eight-hour period they are respiring at almost one and a half times the rate of the non-cleaving eggs, but still only two-thirds that of the fertilized controls. Examination of the eggs from the vessels at the end of the run showed 20 to 30 per cent bottom- swimmers (unicellular) in the case of the non-cleaving eggs (Parth. 5); 100 per cent swimmers of which 30 to 40 per cent were top-swimmers in the case of the cleaving eggs (Parth. 17); and 100 per cent top- swimmers in the case of the fertilized controls.
Fertilized Eggs in Phenylurethane
With the proper concentration of phenylurethane, cleavage may be suppressed without the initial respiratory rate being affected, as
104
ALBERT TYLER AND N. H. HOROWITZ
Warburg (1910) showed on sea urchin eggs. The experiments were repeated for the purpose of covering a longer period. The results of such an experiment with UrecMs eggs are presented in Table IV. The eggs were placed in the solution at 45 minutes after insemination, and the measurements begun 40 minutes later. As the figures in Table IV show, the initial rate of respiration is the same as in the untreated controls. The rate then rises in both, but more slowly in the treated eggs. At the end of the run their respiration is less than three-fourths of the control rate.
It is evident here that the failure of cytoplasmic cleavage is accompanied by a slower rise in respiratory rate. The rate does, however, actually rise. It may be pointed out in this connection that nuclear division goes on in the treated eggs but at a retarded rate.
TABLE IV
Oxygen consumption of phenylurelhane-treated eggs. Eggs placed in 5 X 10~* N phenylurethane after appearance of second polar body (45 minutes after insemina- tion). Measurements begun at 85 minutes after insemination and 40 minutes after immersion in phenylurethane. Values given as cu. mm. Os per hour per mg. N. Temperature = 22.0° C.
|
Hour |
Control |
Control |
Phenylur. |
Phenylur. |
|
1st |
4.61 |
4.64 |
4.86 |
5.04 |
|
2nd |
4.94 |
4.92 |
4.69 |
4.83 |
|
3rd |
5.66 |
5.48 |
5.40 |
6.71 |
|
4th |
6.99 |
7.05 |
6.72 |
5.13 |
|
5th |
9.84 |
10.02 |
9.16 |
10.84 |
|
6th |
13.34 |
13.54 |
10.80 |
9.26 |
|
7th |
16.30 |
16.40 |
12.30 |
14.01 |
|
8th |
18.83 |
18.79 |
13.68 |
14.01 |
|
9th |
21.61 |
21.33 |
15.43 |
15 98 |
Runnstrom (1928) investigated the action of phenyl- and ethyl- urethane on the respiration of sea urchin eggs. The concentrations that he used gave a depression of the initial rate; nevertheless a distinct rise is manifested during the three-hour period of the experi- ments. He notes too that nuclear division proceeds although the respiration during the first hour may be only 35 per cent of that of the control.
DISCUSSION
The curves of Fig. 1 summarize the results. It is readily seen that the parthenogenetic eggs, although starting out at about the same rate as the fertilized eggs, do not give as rapid a rise with time. Tin- t\\o types "t parthenogenetic eggs give approximately the same values (luring the early stages, but later the dividing eggs manifest a
RESPIRATORY RATE AND CLEAVAGE
105
more rapid rise. If only the early period were considered this might be taken to mean that the rise in respiratory rate is not at all correlated with cleavage. However, as pointed out above, the cleavage parthe- nogenetic eggs divide at a retarded rate. Also, it is evident that in normal fertilized eggs no appreciable rise occurs during the early cleavage. It is therefore fairly safe to conclude that the rise in rate is linked with cleavage. But that it is not merely a matter of cyto- plasmic cleavage is clear from the fact that in the non-cleaving parthe- nogenetic eggs and the phenylurethane-treated eggs the respiratory rate does rise with time. Here, as we have seen, nuclear division goes on and it is probably with this factor that the rise is connected.
7 9 II 13
Hours after fertilization
15
17
19
FIG. 1. Rate of oxygen consumption of Urechis eggs. Curves A and A'; fertilized eggs. B and B'; non-cleaving parthenogenetic eggs. C and C'; cleaving parthenogenetic eggs. D; phenylurethane-treated eggs. Temperature 22° C. Values are averages from all of the data. A', B' and C' are not direct continuations of A, B, and C since the eggs had been cultured at 18.2° C. for 11 hours.
Similar experiments have recently been independently performed by Brachet (1938) on Cliaetopterus eggs that " differentiate without cleavage" and he has kindly allowed us to examine his manuscript. The results with Chaetopterus agree very well with those presented here. In addition Brachet has investigated the nucleic acid content and finds it to be much lower in the parthenogenetic (non-cleaving) eggs than in the fertilized controls.
In experiments of this type it is, of course, important to rule out possible direct effects of the chemical agent employed. It is reasonable to assume that there has been no direct effect when the initial rate is unaltered. Such is the case in the experiments reported here with
106 ALBERT TYLER AND N. H. HOROWITZ
the exception of the cleaving parthenogenetic eggs (Table II) which showed a significantly high initial respiratory rate. This is very likely due to the ammonia treatment and from the subsequent values we might assume that the effect passes off. In any event, the differ- ence would have to be greater and in the opposite direction to seriously affect the conclusion.
As was pointed out in the introduction, we might expect no rise in respiration if cleavage were blocked by some means that does not alter the absolute respiratory rate at the particular stage. But we are not dealing here with such ideal cases. In these experiments cytoplasmic division fails (non-cleaving parthenogenetic eggs) or is prevented (phenylurethane-treated eggs) while nuclear division pro- ceeds at a retarded rate; or both cytoplasmic and nuclear division proceed at a retarded rate (cleaving parthenogenetic eggs). It is not surprising, therefore, to find that the respiratory rate does increase in these cases. That the parthenogenetic show a slower rise than the fertilized eggs is consistent with their slower development, as is also the difference between the phenylurethane-treated and the control eggs. The difference between the two types of parthenogenetic eggs may likewise be interpreted as due to differences in the rate at which comparable stages of development are reached.
SUMMARY
1. Artificially activated eggs of Urechis respire at the same initial rate as do normally fertilized eggs.
2. The rate of respiration rises with time in the artificially activated eggs, but at a much slower rate than in the fertilized eggs.
3. The increase in respiratory rate with time is greater with cleaving than with non-cleaving parthenogenetic eggs.
4. Fertilized eggs in which cytoplasmic cleavage is inhibited and nuclear division retarded by means of phenylurethane give a retarded rise in respiratory rate, although the initial rate is the same as the control rate.
5. It is concluded that the delayed rise in respiration is linked with the slower development in all these cases.
LITERATURE CITED
BRACHET, J., 1938. The oxygen consumption of artificially activated and fertilized
Chaetopterus eggs. Biol. Bull., 74: 93. HIKAIWA, J. K., AND T. KAWAMURA, 1936. Relation between maturation division
and cleavage in artificially activated eggs of Urechis unicinctus (von
Drasche). Biol. Bull., 70: 344. LII.LIE, F. R., 1902. Differentiation without cleavage in the egg of tin- annelid
Chaetopterus pergamentaceus. Arch. Entw.-mech., 14: 477.
RESPIRATORY RATE AND CLEAVAGE 107
LOEB, JM AND H. WASTENEYS, 1913. The influence of hypertonic solution upon the
rate of oxidations in fertilized and unfertilized eggs. Jour. Biol. Chem.,
14: 469.
NEEDHAM, J., 1931. Chemical Embryology. The MacMillan Company, New York. RUNNSTROM, J., 1928. Struktur und Atmung bei der Entwicklungserregung des
Seeigeleies. Ada. Zool., 9: 445. TYLER, A., 1931. The production of normal embryos by artificial parthenogenesis
in the echiuroid, Urechis. Biol. Bull., 60: 187. TYLER, A., 1936. On the energetics of differentiation. IV. Comparison of the rates
of oxygen consumption and of development at different temperatures of
eggs of some marine animals. Biol. Bull., 71: 82. TYLER, A., AND H. BAUER, 1937. Polar body extrusion and cleavage in artificially
activated eggs of Urechis caupo. Biol. Bull., 73: 164. WARBURG, O., 1910. Uber die Oxydationen in lebenden Zellen nach Versuchen am
Seeigelei. Zeitschr.f. physiol. Chem., 66: 305.
i:\TKklMI.\Tso\ LIGIA 1\ BERMUDA \'. FURTHER EFFECTS OF SALTS AND OF HEAVY SEA WATER
T. CUNLIFFE BARNES
(From tJte Osborn Zoological Laboratory, Yale University, and tlie Bermuda Biological
Station)
Four previous papers of this series (Barnes, 1932, 1934, 1935, 1936) have described aspects of the behavior, salt requirements and thermal range of Ligia baudiniana, an interesting isopod which is invading the land through the intertidal zone. The present report deals with fur- ther experiments on the capillary mechanism conveying sea water to the gills, reactions to filter paper saturated with diluted sea water, the protective action of Ca in hypotonic sea water and the combined effect of high temperature and heavy water.
HABITS
It was reported previously that the release of young from the brood pouch was observed only in specimens kept submerged in sea water, but during the past summer this phenomenon occurred in a few cases in females in air over filter paper moistened with sea water. On the other hand, the molting process has occurred only in air. No adult specimens have been taken in the sea, but in rare cases an isopod will enter sea water in a terrarium to feed.
Ligia is provided with a capillary mechanism for keeping the gills moist without entering the sea. The first paper of this series stated that the uropods and spines are lowered into the water or onto a water film and the sea water then rises by this capillary path to the gills. However, Mr. M. D. Burkenroad has drawn my attention to a more important capillary conduit by which the water rises between the sixth and seventh legs to the gills and is then propelled by the gills down the uropods and spines (Fig. 1, A). The sixth and seventh leg on one or both sides may be used. Frequently the animal slides one leg over the other alternately at the start. The drainage down the uropods was described in a former paper (Barnes, 1935) in experiments in which sea water was dropped on the animals. The flow was followed in the present experiments by the addition of stains, fine particles or bubbles to the sea water. If the uropods are lowered when the last pair of legs are not drawn together, the first movement of the water is ii] i I lie capillary conduit of the uropods as originally described (which
108
EFFECTS OF MODIFIED SEA WATER ON LIGIA
109
may be observed by the movement of particles in the sea water). When the isopod is totally immersed, the currents produced by the beating of the gills in the surrounding sea water (indicated by the move- ment of particles) follow the same direction as the capillary circuit involving the rise of water between the last pair of legs (Fig. 1, B).
The sixth or seventh leg was removed in some specimens to modify the method of obtaining water. In the absence of the seventh leg a specimen standing on moist filter paper usually takes up the position
FIG. 1. Methods of moistening the gills in Ligia. (In these sketches the an- tennae and body segmentation are omitted.)
A. Specimen in air on filter paper saturated with distilled water. Water rises between the sixth and seventh leg by capillarity and is drained down between the uropods as indicated by arrows.
B. Specimen immersed in sea water. The arrows indicate the currents main- tained by the beating of the gills.
C. Specimen with seventh leg removed standing on filter paper saturated with sea water. Sixth leg and uropods are in position for capillary circuit.
D. Side view of specimen with sixth leg removed standing on film of sea water (stippled).
E. Side view of specimen with sixth leg removed with abdomen lowered in film of sea water.
F. Sixth and seventh legs permanently fused along segment indicated by dotted line (found on left side of one specimen). These partially fused legs functioned normally as a capillary channel.
for the capillary circuit via the legs (Fig. 1, C). It feels about franti- cally with the sixth leg for the missing seventh leg to complete the capillary channel. No attempt is made to use the fifth leg. Similarly if the sixth leg is removed, the conduit is impossible. Some specimens "squat" on the water film and the water rises by capillarity on the whole undersurface of the body and gills (Fig. I, D). Other specimens with the sixth leg missing lower the abdomen and uropods and thus
110 T. CUNLIFFE BARNES
secure water by capillarity (Fig. 1, £). It is clear that there are several methods of wetting the gills. Indeed, many specimens are found which have lost the uropods and spines. These simply lower the abdomen on a damp substratum.
An interesting specimen was found in which the sixth and seventh leg were permanently fixed in the capillarity posture owing to the fusion of the last segments (Fig. 1, F). While walking, the double leg moved alternately with the seventh leg on the opposite side, but sometimes dragged. The joined legs functioned normally as a channel above the point of fusion.
REACTION TO FILTER PAPER SATURATED WITH DILUTED
SEA WATER
Large filter papers (diameter 25") were cut in two, one half satu- rated with distilled water, the other with diluted sea water and placed
TABLE I
Reaction of Ligia to Filter Paper Saturated with Diluted Sea Water (The animals were tested in groups of four)
|
Total number of |
||
|
Treatment of each half of paper |
isopods found on |
Ratio |
|
each half |
||
|
Sea water vs. distilled water |
69 : 123 |
1 : 1.78 |
|
75 per cent sea water vs. distilled water . |
27 : 73 |
1 : 2.74 |
|
50 per cent sea water vs. distilled water . |
93 : 96 |
1 : 1.03 |
|
25 per cent sea water vs. distilled water . |
109 : 73 |
1 : 4.9 |
|
10 per cent sea water vs. distilled water . |
27 : 21 |
1.28 : 1 |
in a covered flat dish in a photographic darkroom having a light directly above the center of the dish. Four isopods (previously kept on sea- weed moistened with sea water) were placed in the dish and their distribution on the two halves was observed at ten to fifteen-minute intervals over a period of one to two hours. The dish was rotated 90° after each reading to eliminate any unsuspected source of orienta- tion. The animals showed a distinct tendency to collect on filter paper saturated with distilled water when the other half was moistened with sea water (Table I, and Barnes, 1935). It was found that this aversion was also shown to 75 per cent sea water, but not to dilutions of 50 per cent and below (Table I).
The tendency to avoid sea water was most pronounced during the first observations in a given experiment and after an hour or more, when the paper containing distilled water was becoming dry, most specimens collected on the sea water side. Thus the ratio of specimens on the sea water side to those on the distilled water side was 1 : 3.3
EFFECTS OF MODIFIED SEA WATER ON LIGIA
111
for all the first readings and 2.3 : 1 for the last observation made after an average time interval of one and one half hours.
THE PROTECTIVE ACTION OF CALCIUM IN HYPOTONIC SEA WATER
Ligia survives for only about seven hours in 100 cc. of 25 per cent sea water, but as with many other forms, the addition of calcium protects the organism from hypotonic media, apparently by decreasing permeability. As in all the experiments with solutions, individual specimens were tested in 100 cc. of solution in finger bowls. As will be seen from Table II, there is a threshold for the Ca effect (at about
TABLE II
Longevity of Ligia in 25 per cent Sea Water Containing Added
|
cc. f M CaCh |
||||
|
added to 1 liter |
Average |
Maximum |
Coefficients of |
Number of |
|
25 per cent |
longevity |
longevity |
variation |
specimens |
|
sea water |
||||
|
hours |
hours |
|||
|
0 |
7 ± 0.33 |
15 |
4.7 |
33 |
|
2 |
6.5 ±0.20 |
11 |
3.0 |
30 |
|
5 |
7.6 ± 0.37 |
17 |
4.8 |
58 |
|
12 |
24.6 ± 3.66 |
119 |
14.8 |
30 |
|
15 |
24.1 ±3.70 |
144 |
15.3 |
20 |
|
20 |
51.5 ± 7.98 |
120 |
15.5 |
20 |
|
25 |
37.8 ± 5.09 |
168 |
13.5 |
30 |
|
50 |
38.5 ± 3.85 |
180 |
10.0 |
40 |
12 cc. | M CaCl2 added to a liter of 25 per cent sea water). It is also apparent that the addition of Ca above a critical quantity (20 cc. f M CaCl2 added to a liter of 25 per cent sea water) has little additional protective action. It is interesting to note (see Fig. 2) that the average survival in calcified 25 per cent sea water approaches the aver- age longevity in 100 cc. of natural sea water.
THE REVERSIBILITY OF THE CALCIUM EFFECT
Experiments were carried out to see if the protective action of calcium involves an irreversible chemical process. If so, preliminary treatment in solutions of high calcium content should lengthen the survival of isopods subsequently immersed in hypotonic sea water. However, it was found that preliminary immersion in calcified sea water from half an hour to over a day had little if any effect on the subsequent longevity of the treated specimens in 25 per cent sea water (see Table III). Likewise, preliminary treatment with sea water having a high content of sodium has no significant effect on subsequent action of hypotonic sea water. In all these tests many specimens
112
T. CUXLIFFE BARXES
died during the preliminary treatment so that those tested in the 25 per cent sea water represented a selected population capable of with- standing the submerged state. It must be remembered, however, that the survival of Li^ia in any liquid medium is limited by unknown factors so that these selected specimens were already weakened by submersion. That these two factors balanced each otht-r is indicated by the similarity of the average longevity in 25 per cent sea water of untreated and treated specimens. Thus the 152 isopods exposed to
50
40
30
20
10
10
20
30
40
50
Ki<;. 2. Length of life of Ligia in 100 cc. of 25 per cent sea water with increasing amounts of CaCl2.
Onlinates, average life in hours. Abscissa1, cc. of jj M CaCI2 added to 1,000 cc. of 25 per cent sea water. The graph represents 228 tests (see Table II). Lower dotted line is the expected survival in 25 per cent sea water. Upper dotted line is the average survival in natural sea water. The vertical lines are proportional to the probable errors. The sea water average represents 265 specimens and the 25 per cent sea water average, 63 specimens. These are cumulative totals for five seasons' work. In all cases an individual specimen was tested in 100 cc. of solution.
previous solutions survived 8.9 hours in 25 per cent sea water compared to the usual average of about seven hours in this medium for freshly collected individuals. In all the tests the specimens were washed by rapid immersion in distilled water after treatment in tin- calcium solutions, but this had no effect on the subsequent survival in 25 per cent sea water. Thus controls treated with ordinary sea water and
EFFECTS OF MODIFIED SEA WATER ON LIGIA
113
then washed in distilled water lived (he usual interval of about 7 hours in 25 per cent sea water.
Experiments in which the isopods were tested in distilled water after treatment with CaCl2 give similar results (see Table IV). After exposure to f M CaCl2 the average survival was three hours in distilled water compared to 2.8 hours for untreated specimens.
THE COMBINED EFFECTS OF HIGH TEMPERATURE AND HEAVY WATER
It was previously shown that Ligia succumbs rapidly in a small volume of sea water at 38° C. The following experiments were performed to ascertain if sea water containing 99 per cent heavy water would protect the organism against heat death or else potentiate with high temperature by reducing the survival time. In each case 5 cc.
TABLE III
Longevity of Ligia in 25 per cent Sea Water after Treatment in Calcified Sea Water
|
Subsequent |
|||||
|
Average |
average |
Maxi- |
Coefficients |
Number |
|
|
Initial solution |
length of |
longevity |
mum |
of |
of |
|
treatment |
in 25 percent |
longevity |
variation |
specimens |
|
|
sea water |
|||||
|
hours |
hours |
hours |
|||
|
Sea water . . |
12.2 |
6.2 ±0.30 |
13.5 |
4.8 |
34 |
|
50 cc. f M CaCl2 plus |
|||||
|
950 cc. sea water .... |
31.5 |
5.5 ±0.32 |
6.5 |
5.8 |
13 |
|
100 cc. f M CaCl2 plus |
|||||
|
900 cc. sea water .... |
21.1 |
6.5 ±0.33 |
9 |
5.0 |
22 |
|
200 cc. f M CaCl2 plus |
|||||
|
800 cc. sea water .... |
17.6 |
9.2 ±0.38 |
17 |
4.1 |
49 |
|
500 cc. | M CaCl2 plus |
|||||
|
500 cc. sea water .... |
0.5 |
10.5±0.39 |
18 |
3.7 |
40 |
|
400 cc. | M NaCl plus |
|||||
|
600 cc. sea water .... |
25.3 |
9.4±0.79 |
25 |
8.4 |
28 |
of sea water were evaporated and the salts redissolved in ordinary distilled water or in 99 per cent heavy water. It was found that death ensued twice as rapidly in heavy sea water at 38° C. The survival times in 5-cc. samples were 23 minutes for H2O sea water and eleven minutes for D2O sea water (see Table V) . Owing to the scarcity of heavy water, several specimens were tested in the same sample, but no appreciable difference was observed in the survival times of the first and last isopods treated.
DISCUSSION
The tendency of the isopods to collect on filter paper treated with distilled water rather than on sea water paper is of interest in connec- tion with the animals' aversion for the sea. The reaction appears to
114
T. Ct'XLIFFE BARXES
TAHLE IV
Longevity of Ligia in Distilled Water after Treatment with Calcium
|
Subsequent |
|||||
|
Average |
longevity |
Maxi- |
Coefficients |
Number |
|
|
Initial solution |
length of |
in |
mum |
of |
of |
|
treatment |
distilled |
longevity |
variation |
specimens |
|
|
water |
|||||
|
minutes |
hours |
hours |
|||
|
Xone |
0 |
2. 8 ±0.2 5 |
5.2 |
8.9 |
22 |
|
500 cc. § M CaCI2 plus |
|||||
|
500 cc. sea water .... |
33 |
4.5 ±0.15 |
6 |
3.3 |
20 |
|
| M CaCU |
12 |
3±0.20 |
4 |
6.7 |
12 |
be determined by the salt content of sea water since dilution of 50 per cent destroys the effect. It is possible that the salts on the paper stimulate the isopods to greater movement which would cause them to collect on the salt-free side. Gunn (1937) has described a hygro- kinetic effect in Porcellio scaber whereby the greater activity of animals in dry air causes them to collect in moist locations, but this mechanism is probably not responsible for the reaction of Ligia described above. As was shown in a previous paper (Barnes, 1935), specimens will collect on dry filter paper when the other side of the dish contains filter paper moistened with distilled water. This refers to moist specimens. It is not known what part the flushing mechanism of the gills plays in
TABLE V
Longevity of Ligia in 5 cc. of Ordinary and Heavy Sea Water at High Temperature
|
Medium |
Tempera- ture |
Average longevity |
Maximum longevity |
Coefficients of variation |
Number of specimens |
|
Sea water |
°C. 38 |
seconds 1383 ±45.1 |
seconds 2005 |
3.26 |
27 |
|
Sea water containing 99 per cent D2O |
38 |
660±57.0 |
1140 |
8.63 |
11 |
these reactions. Bateman (1933) found that salts become slightly concentrated in the blood of specimens kept in air and it is possible that flushing the gills, especially with distilled water, enables the animal to get rid of salts concentrated by evaporation. On the other hand, specimens which have been immersed in distilled water also show t he- aversion for paper soaked in sea water (Barnes, 1935).
The well-known action of ('a as a factor enabling organisms to withstand hypotonic solutions (for reference cf. Barnes, 1934) is strikingly illustrated by Ligia, which survives almost as long in 100 cc. of 25 per cent sea water of approximately 0.015 M CaClo content as in 100 per cent of natural sea water. To the list of favorable artificial
EFFECTS OF MODIFIED SEA WATER ON LIGIA H5
sea water solutions for Ligia described in previous papers may be added this new hypotonic mixture which is equivalent to sea water in which the Na, K and Mg content has been reduced to one fourth and the Ca content raised slightly. However, the maximum longevity in this medium is far short of the 297 hours observed in natural sea water. The very long survival of occasional specimens in sea water suggests that other factors besides salt effects are involved, such as the amount of food in the gut, the nature of the previous environment, the oxygen content, or the necessity for molting (which apparently does not occur in sea water). The high summer temperature of approximately 27° C. must also be considered.
The failure of preliminary immersion in sea water rich in calcium to protect the organism against subsequent exposure to hypotonic sea water indicates that Ca forms a loose, rapidly reversible combination with material in the plasma membrane. The existence of a threshold concentration and of a limited range in which the effect of Ca is pro- portional to the concentration suggests that a surface reaction is involved.
The rapid lethal action of 5 cc. of sea water at 38°, in which the ordinary water has been replaced by heavy water, was an unexpected result. The lower energy content of heavy water, which under certain conditions might be expected to protect an organism from high temperatures, was overbalanced by the well known toxic action of high concentrations of deuterium. Barbour (1937) has recently dis- cussed the toxicology of heavy water. In larger volumes of sea water or even fresh water Ligia withstands high temperatures for a much longer period, probably on account of the greater quantity of oxygen present. The 5-cc. samples used in the tests were exposed to the air in a Petri dish so that the oxygen content of both heavy and ordinary sea water was probably the same.
SUMMARY
1. When presented with a "choice" between filter paper moistened with sea water or with distilled water, freshly caught specimens of Ligia tend to collect on the latter. Dilution of the sea water destroys this effect.
2. The survival of Ligia immersed in 25 per cent sea water with added calcium approaches the longevity in natural sea water.
3. Preliminary exposure of solutions of high calcium content does not protect Ligia from subsequent immersion in hypotonic sea water.
4. The lethal action of 5 cc. of sea water at 38° C. is enhanced by 99 per cent heavy water.
116 T. CUNLIFFE BARNES
LITERATURE CITED
BARBOUR, H. G., 1937. The basis of the pharmocological action of heavy water in
mammals. Yale Jour. Biol. and Med. 9: 551. BARNES, T. C., 1932. Salt requirements and space orientation of the littoral isopod
Ligia in Bermuda. Biol. Bull., 63: 496. BARNES, T. C., 1934. Further observations on the salt requirements of Ligia in
Bermuda. Biol. Bull., 66: 124. BARNES, T. C., 1935. Salt requirements and orientation of Ligia in Bermuda. III.
Biol. Bull., 69: 259. BARNES, T. C., 1936. Experiments on Ligia in Bermuda. IV. The effects of
heavy water and temperature. Biol. Bull., 70: 109. BATEMAN, J. B., 1933. Osmotic and ionic regulation in the shore crab, Carcinus
maenas, with notes on the blood concentrations of Gammarus locusta and
Ligia oceanica. Brit. Jour. Exper. Biol., 10: 355. GUNN, D. L., 1937. The humidity reactions of the wood-louse, Porcellio scaber.
Brit. Jour. Exper. Biol., 14: 178.
SOME ASPECTS OF NORMAL DEVELOPMENT IN THE COLONIAL CILIATE ZOOTHAMNIUM ALTERNANS
F. M. SUMMERS
(From the Marine Biological Laboratory, Woods Hole, and Bard College, Columbia University)
INTRODUCTION
This is the first of a series of inquiries into some of the problems of organization and regulation in a colonial protozoan of a rather special type, Zoothamnium alternans, whose cells collectively possess in some degree many of the attributes of an individual organism. In order to fix some standard by which to judge experimental results it was first necessary to review in detail the characteristic features of normal development in a large number of cases.
Although Claparede and Lachmann (1858) formulated a scheme of notation for cell lineage in this species, Faure-Fremiet (1930) was the first to make a comprehensive study of its development. His primary concern was the demonstration of cytological mechanisms by which potentialities for growth and ciliospore formation were assorted to cells in characteristic positions on the colony. According to his conclusions this "cytological" factor determines the general pattern of develop- ment. His second objective, a study of time as a limiting growth factor, has no immediate bearing on the subject to be treated here. The quantitative aspects and time relations in development are to be incorporated in a subsequent report.
The data presented here augment in considerable detail the previous studies of Zoothamnium alternans, particularly with regard to the features of development which have an important bearing on the problems of regulation now being investigated.
The work was done at the Marine Biological Laboratory, Woods Hole, Massachusetts, during the summer months of 1935-6.
It is a privilege to acknowledge assistance from those whose friendly interest has sustained the author's enthusiasm over many of the dull periods of routine work.
MATERIAL
The basal portions of mature hydroids (Pennaria and Tubularia} and Bryozoa (Bugula} from many localities in Woods Hole harbor great numbers of attached colonies of Zoothamnium alternans. The
117
118 F. M. SUMMERS
hydroid and hryozoan colonies are clean and free from the ciliates until about June 15. From then on throughout the summer the proto- zoa are very abundant, although their condition varies with particular habitats and with the water temperature. Grave (1933) found that the hydroids of this region are particularly healthy and free from bacterial growths, debris, etc. for the early spring months but that in mid-summer the contaminating organisms affect them adversely. This applies equally to the associated protozoa, for those obtainable in late July and August are commonly infested with parasites of several classes. The most annoying and destructive organisms are bacteria, filamentous algae, and an intra-cellular parasite which has been identified as a suctorian (Acineta) by Faure-Fremiet (1930); others of less consequence are small Crustacea, diatoms, and other ciliates, e.g. Amphileptits and Cothurnia.
METHODS
The procedure for maintaining selected colonies in aquaria is an elaboration of the technique employed by Faure-Fremiet (1930) for his study of growth in Zoothamnium alternans. Hydroids bearing the protozoa were cleaned, washed, and packed into large glass vessels of running sea water. Ruled and numbered slides were placed face down on the infected hydroid material and left for about 12 hours. They were then removed and washed by passing under a stream of running water. The precautions taken at these stages to remove adhering scum, debris, algae, other protozoa, and minute Crustacea, determined to a great extent the age to which the colonies could be grown .
Some of the mature asexual migratory cells, the ciliospores (Wesen- berg-Lund, 1925), liberated from the "wild" colonies during the 12- hour interval, attached themselves to the ruled surface of the numbered slides. Slides were examined in Petri dishes, the positions of the recently affixed ciliospores recorded, and then transferred from the Petri dish to correspondingly numbered slots (vertical) in wooden racks, immersed in a 15-gallon aquarium. Cultural conditions were very much improved by the use of filtered sea water. Two 2-liter aspirator bottles packed with sand and stoppered with glass wool proved to be adequate for the purpose. When cleaned daily their combined outputs averaged about 50 gallons per hour.
Routine counts were made with a 4 mm. dry objective, \\hen optical sections of the barely immersed colonies were not clear, the water level in the Petri dish was adjusted by a slight turn of the dish on the slanted microscope stage until the surface film flattened the
NORMAL DEVELOPMENT IN ZOOTHAMNIUM ALTERNANS 119
colony. This was done as a last resort, for contact with the surface film elicited rapid contractile responses, frequently resulting in the loss of important zooids.
7MJ9
FIG. 1. A relatively mature colony of Zoathamnium alternans showing the characteristic arrangement of the zooids and branches. Axial microzooids on branches E, G, L, and N are differentiating into ciliospores; one of them at E is in an advanced stage of metamorphosis, almost ready to break away from the parent colony. Approximately X 100.
GENERAL FORM OF THE MATURE COLONY
The tapering frond-like configuration of the older expanded colonies is shown in Fig. 1. Its branches project obliquely from the per-
120 F. M. SUMMERS
pendicular axis. They lie in the same plane but alternate on right and left sides at successive nodes. The zooids of the branches occur singly for the most part, likewise alternating in position along the branch axis. While the branches of this species usually consist only of a single axis, in rare cases the zooids may divide one or more times to produce branches of the second order.
The expanded colony is essentially asymmetrical, i.e. characterized by three heteropolar axes. The primary axis is firmly cemented to the substrate at the basal end and bears the actively dividing terminal macrozooid (TM.) at the apex. The lateral axis is heteropolar by virtue of the alternating branches, each one terminating in a single cell, the dominant branch zooid. "Anterior" and "posterior" sur- faces (Faure-Fremiet, 1930) are defined by the curvature of the branches as well as by the position of the zooids: the branches curve anteriorly while the zooids are inclined posteriorly (toward the convex surface). A sample count showed that the first branch occurred nine times on the left side to four on the right. Both dexio- and leiotropic colonies were reported in another species, Zoothamnium arbuscula, by Furssenko (1929).
Beginning at a point about mid-way between the basal (attached) end of the stalk and the first branch, a heavy contractile central cordon, or "neuro-muscular" cord, within the stalk substance ramifies through- out the colony (Fig. 1). Its branches terminate distally at the basal ends of the zooids. It is thus continuous from branch to branch and from cell to cell. More will be said about the importance of this structure in a later section.
NORMAL DEVELOPMENT
The scheme used to designate the lineage of the zooids is a modified form of Faure-Fremiet's adaptation of the original rule formulated by Claparede and Lachmann (1858-60), who were the first to note devel- opment in this species according to a definitely determined pattern.
The mode of propagation in Zoothamnium alternans is almost exclusively asexual. Axillary branch zooids at various levels in larger colonies are transformed into migratory ciliospores. These liberated motile individuals are comparatively large cells, metamorphosed zooids, whose sub-conical bodies flatten antero-posteriorly into cells having a thickened bi-convex lens-like appearance. Their rapid swimming or creeping movements are effected by means of a heavy equatorial girdlet of cilia. The oral disc and peristomal cilia are retracted and probably undergo reorganization during the motile phase.
NORMAL DEVELOPMENT IN ZOOTHAMNIUM ALTERNANS 121
After a migratory existence of several hours duration the ciliospores become relatively quiescent, hovering about within a limited radius with aboral end in contact with the substrate, finally fixing themselves to the surface by means of a scopula. The secretion of the stalk begins ; the girdlet of cilia becomes inactive and disappears as the peduncle elongates. The lens-shaped ciliospore then begins to assume the appearance of a very large sub-conical zooid with retracted adoral zone. The first section of the peduncle secreted is a fairly heavy, non- contractile cylinder with thickened cuticle and homogeneous hyaline medullary substance. At some point between 200-30(V the secretory activity of the ciliospore is altered: the continuity of the secretory process is interrupted, the basal end of the cell increases in diameter, and the formation of the neuro-muscular cord begins. This node marks the lower limit of stalk contractility. The contractile cord is an integral part of the stalk structure from this point onward.
The division of the initial individual occurs approximately 15 hours after attachment. The first division furrow defines the median plane of the future colony. Succeeding furrows bear a definite and constant relation to the first, in such a manner that the daughter cells occur first on one side of the mid-line then on the other. The two daughter cells resulting from the first division are always markedly unequal in size (Fig. 2). The large daughter remains apical (terminal) in position.
FIG. 2. Developing colony at the 2-cell stage. Drawn from a living specimen in which the macronuclei were faintly visible. X 250.
In the current terminology it is designated as the terminal macrozooid No. 1. The small lateral daughter, the first cell of the first branch, is characterized as the median microzooid A.
Dichotomous branching of the stalk occurs when the secretory basal portion of the cell constricts during the fission process. With the onset of secretory activity (i.e. stalk formation) the sister cells gradually move away from the point of origin. Material secreted by the terminal macrozooid prolongs the axial stalk. The branch cell produces a somewhat smaller stalk extending laterally from the junction.
122 R M. SUMMERS
The apical cell (TM. No. 1) divides again at the second node in approximately 8 hours, producing an apical cell of the second genera- tion (TM. No. 2) and a median microzooid B. This time, however, the branch cell lies on the opposite side of the axis from the previously produced microzooid A. Repeated divisions of the apical cell extend the principal axis of the colony, each time similarly producing a ter- minal macrozooid of the succeeding generation (3, 4, 5, . . .) and an alternating branch zooid of a higher level (C, D, E, . . .). Tin- position of successive branch microzooids is strictly alternative. Colonies with as many as 25 branches were frequently observed on the culture slides but it is doubtful that this represents maximum develop- ment. One colony produced 33 branches before succumbing to external parasites.
Following the first division of the apical cell there is a gradual reduction in the size inequality of the resulting daughters until, after 8 to 10 generations of axial development, or in some cases even fewer, their volumes immediately after a mitotic period are approximately equal. For experimental purposes it is essential to distinguish between recently formed daughters of later generations along the main axis. To this end it is necessary to refer back to the previous generation. If the left-hand cell of that generation is differentiating into a branch cell, then the right-hand position of the undifferentiated cell at the new mode may be taken as assurance of its presumptive branch rela- tions. Likewise position is tin- chief diagnostic criterion for the identification of undifferentiated branch zooids.
The generalized pattern of branch development given below con- cerns only the position of the cells produced, for as will be seen later, the descendants of the median microzooids are not always equivalent in their prospective values. The symbol X (or .v) is used to denote branch generations in general (Fig. 3).
Lateral extension of the stalk bearing the median microzooid A" continues until the cell divides. Microzooid A' gives rise to two struc- turally similar cells, microzooids 1AT and l.v. Microzooid 1A" occupies a median position with respect to the main axis of the colony, whereas \x is the lateral member. Actually l.v immediately assumes the terminal position on the branch axis, thus constituting the growing point of the branch. Marked differential behavior now characterizes the two sisters: 1A' lays down a short segment of stalk and then remains quiescent for some time ; on the other hand, l.r actively extends the branch axis, then divides again within a few hours. Microzooid l.v then produces 2*1 (median) and 2x2 (lateral). Cell 2.v2 remains behind as a common nutritive branch zooid while 2.Y1 assumes the terminal
NORMAL DEVELOPMENT IN ZOOTHAMNIUM ALTERNANS 123
position and continues as before, giving origin to 3.x:1 (lateral) and 3jc2 (median) ; 3.V1 to 4*1 and 4jt2, etc. Each time the alternate secondary daughter is left behind as a common zooid (Fig. 1).
Further mitotic activity in the common branch zooids 2x2, 3xz, 4#2, etc. was not characteristic of the normal colonies whose histories were charted. About 2 per cent of the total number of colonies under surveillance showed a tendency toward precocious development. In these cases the repeated divisions of all branch zooids led to such complex formations that the lineage soon became impossible to follow.
Conversely, microzooid \X on branches up to about level J com- monly divides once, forming \Xi (median) and IX2 (lateral). Its mitotic activity on more distal branches is infrequent. On several
Presumptive Ciliospores
FIG. 3. Schematic representation of branch development.
branches of each colony the cell IX or its descendants metamorphose into the migratory ciliospores which break away from the parent colony, leaving behind a short, stubby peduncle. Occasionally IX produces a second or sometimes a third generation thereby augmenting the number of propagatory zooids.
DIFFERENTIATION
The distribution of the heteromorphic zooids to be described is a generalized account based upon a tabular evaluation of 200 cases of normal and experimentally cut colonies (proximal portions only), many of which were lost before complete development. Inclusion of the basal parts of operated colonies in the tabulations seems to be justified since statistical summaries of both normal and operated
124 [-. M. SUMMERS
colonies indicate no consistent variation as regards either rate of development or position of the differentiating zooids.
1 hiring the life history of a colony six types of zooids may be formed although not all of them are apt to be present at any one time.
1 . The common nutritive microzooids constitute the majority of the colony cells. They are the trumpet-shaped individuals characteristic of the genus. Notwithstanding the fact that most of them never assume any other form, they may be regarded as generalized com- ponents of prime colonies because (a) under undisturbed laboratory conditions some of them at random positions along the branch axis metamorphose into microgamonts, and (b) experiments may be per- formed in which variously situated common zooids may be induced to differentiate into actively reproducing apical cells. Their capacity to differentiate into another type under imposed conditions gradually diminishes in time so that in older colonies the "vegetative" function becomes fixed. They can then be considered as somewhat specialized zooids. The shape of zooids that have persisted for relatively long periods of time without division are decidedly more slender and elongate than the "average" zooids. This appears to be a modifica- tion of form rather than an alteration of volume. The extension of the protoplasmic body as well as its peduncle also appears to be gradual and continuous subsequent to the post-mitotic growth period.
2. Each colony bears a single terminal macrozooid (TM.} at the apex of the principal axis (Figs. 1 and 2). In the earlier stages of colony formation three structural peculiarities clearly distinguish it from the common zooids. It is considerably larger than any of the common types for the first ten or more axial generations, the body is characteristically flexed, and numerous annular striations in the pellicle are very pronounced. After ten generations or so its form individuality progressively diminishes. At levels beyond the twentieth branch only its position and generative activity are distinctive.
3. The terminal branch zooids (Lv, 2.v', 3.r', . . .) are almost diminu- tive replicas of the terminal macrozooid. For the first few generations they are identified shortly after dividing by fairly marked pellicular annulations and a moderate anterior flexure of the tapering body. Near the close of branch development their patent features are obscure. Standard criteria of position and, to some extent, proclivity toward more rapid stalk formation and division must be re-lied upon to dis- tinguish them from their common neighbors, the secondary daughters. Finally, at the terminus of a fully formed branch the last pair of zooids are usually indistinguishable. This point represents about four generations on branch A, ten at intermediate levels (e.g. //), and progressively fewer after that (Fig. 1).
NORMAL DEVELOPMENT IN ZOOTHAMNIUM ALTERNANS 125
4. The most striking of the heteromorphic zooids are the immature ciliospores, or asexual propagative cells which represent differentiated microzooids of the order \X or descendants. It is important for later discussion to note that the function of ciliospore formation is normally but not invariably restricted to these microzooids. They have been observed to arise from microzooids 5d2 and 6d~ in one instance, and K)//1 and 10//2 in another. Faure-Fremiet (1930) found that the terminal macrozooid was sometimes transformed into a migrating ciliospore but, unlike the usual ciliospore-forming cells, was unaccom- panied by an endomictic reorganization.
Prior to ciliospore formation microzooid IX undergoes a profound growth in size. Very accentuated concentrically disposed pellicular furrows appear during the growth process. It is impossible to predict whether or not it will divide, for in a small percentage of the cases it is transformed directly into a motile ciliospore thereby terminating the lineage of that cell on the colony. Division of IX in from 20 to 70 hours after derivation from the median microzooid X is the rule, however. It is noteworthy, perhaps, that fission may occur after the onset of differentiation. The increased volume at this time is con- siderably in excess of the pre-divisional growth of the other zooids. The two descendants, IX1 (median) and l^"2 (lateral), are seldom equivalent in their size relations; the lateral daughter is usually the larger and the first to metamorphose.
Complete differentiation into motile cells already characterized involves the appearance of a girdle of long, closely set cilia about the equatorial region, the accentuation of the annulations, introversion of the oral disc, and differential growth in the plane of the diameter, eventually flattening the zooids into disc-shaped bodies. It is believed that they are able to feed to some extent when the oral disc is intro- verted and the adoral zone folded. The gullet remains open and the undulating membrane within persists in its activity. The final metamorphosis of IX1 may be deferred until a second generation of potential ciliospores (IX11 and l^12) are produced. The origin of at least four ciliospores from some axial microzooids has been observed. Whether or not they represent three generations of alternate develop- ment or two generations formed dichotomously is still an open question.
The production of ciliospores is restricted to a limited number of loci on any given colony. Commonly there are but three or four such loci on one colony, although one protocol shows as many as seven, generalized colony as visualized from tabulated data bears florescences at levels D-K-P-T-AA. They have been recorded for all nodes along
126 F. M. SUMMERS
the axis between B to IT inclusive, hut never at consecutive nodes in any one case. It should also be stated that 1A" on many branches divides without giving rise to ciliospores. Its descendants persist as common microzooids.
5. Microgamonts (microgonidia, Wesenberg-Lund, 1925; micro- conjugants, Fursscnko, 1()29) develop from otherwise visibly undiffer- entiated microzooids at the terminal or sub-terminal position along branch axes. The observed time interval between the formation of the microzooids and their complete transformation into mature, liberated swimmers varies between the limits of six to thirty-six hours. The first sign of approaching metamorphosis is the appearance of a faint furrow about the equatorial region of the tapered body. The zooid is thus marked into a larger peristomal portion and a smaller basal portion. Cilia gradually grow out from the body at the furrow. At first very short, they soon develop into a broad girdle whose move- ments simulate those of a slowly waving undulating membrane (Fig. 4). Meanwhile the diameter of the cell increases. The oral half
FIG. 4. Differentiating microgamont about 5 hours before the completion of metamorphosis. X 250.
containing the retracted peristomal disc is transformed into the slightly flat, sub-conical oral end of the swimmer. The posterior portion is further reduced to a very slightly convex aboral surface at whose center the stalk is still attached. Its struggling movements gradually accelerate until at full maturity it breaks away from the parent colony. At full term the microgamonts resemble ciliospores in their eccentric body outline, ciliature, and general movements. They are, however, not perceptibly larger than the common zooids, show but faint indica- tions of annulations, and swim more rapidly.
While cytological studies were not attempted, it is definitely es- tablished that these are the migratory members of the copulatory series. When liberated some have been observed to swim rapidly back and forth in an arc about the parent colony, occasionally coming to rest on some zooid, or crawling about on the branches and over the individuals in various positions, and eventually swimming away as t hough its relatives were not receptive of its attentions. The duration of the migratory phase was not determined. One was observed to settle upon terminal macrozooid No. 3 of a young colony possessing nine cells in all. Two and one-half hours later the aboral end of the
NORMAL DEVELOPMENT IN ZOOTHAMNIUM ALTERNANS 127
migrant was firmly fused with the lateral surface of the apical cell. By that time the ciliary crown of the swimmer had been resorbed.
As regards origin, it is frequently difficult to determine if micro- gamonts arise from prospective terminal or sub-terminal branch microzooids because of the facility with which the prospective values of very recently produced daughter cells may be altered. For example, if the terminal cell (e.g. 2x1} is accidentally lost, the sub-terminal cell (2.r2) immediately assumes the terminal functions. If loss occurs soon after fission, when both daughters are superficially alike, the stalk remnant left behind at the node occurs in the alternate position, whereas the cell which normally would have been the alternate (2^2) becomes the growing point. The tendency of the terminal branch cell to differentiate into a microgamont is conclusively demonstrated by the successive metamorphosis of sister cells. This means that the continued lateral growth of a branch subsequent to the production of a microgamont may represent each time an early (undetected) trans- formation of the secondary zooid into a new terminal cell. The con- fusion is apt to arise because the alternate position of the stubby stalk left by the lost cell or by the migrating microgamont cannot always be diagnosed accurately. Fortunately in many instances the terminal branch cell is well indicated either by position or by symptoms of an approaching division by the time the alternate cell shows signs of becoming a microgamont.
Tabulations show that 85 per cent of the microgamonts observed on colonies whose complete developments were charted, were produced by the four basal branches A, B, C, and D, enumerated in decreasing order of importance. The rest arose from zooids on branches E, F, G, and K. They were formed either by the terminal (-xl) or alternate (-x2) cells as far out on the branches as the sixth generation on A, B, and C (including 5a21 and 5a22), the seventh generation on D, and the eighth generation on F. None were observed to come from the median microzooid X, the axial microzooid IX or its descendants.
The majority of the colonies bore but one or two microgamonts; a few produced as many as nine but not more than three for any one branch. They occurred simultaneously on different branches but successively from the alternate zooids on any given branch. And finally, it is important that they differentiated only from branch zooids at or near the terminal position, never from the more axial (older) microzooids.
6. Macrogamonts were found only in the terminal macrozooid position on the primary axis. Attempts to discover some morphologi- cal or behavioral distinction between terminal macrozooids and pre-
128 R M. SUMMERS
conjugant macrogamonts were negative, but this possibility is not precluded since of the three or four hundred colonies inspected, none may have happened to have been preparing for conjugation. The presence of an attached microgamont constituted the sole criterion for identification.
Conjugants at all levels between the terminal macrozooid positions three to twenty-four inclusive were observed. The youngest colony possessed but nine cells in all, including the conjugant. Two succes- sive conjugants, five generations apart, were recorded for one colony.
The incidence of conjugation reached a maximum during the month of July; before or after that the conjugants were quite rare. Even in July only about 5 per cent of the colonies bore conjugants although more of them were producing microgamonts. Many of the colonies produced no gamonts whatever during the period of development on the slides.
Some of the effects of conjugation upon colonial development will be treated in a later section.
REACTIVE CAPACITIES
The motor responses of Zoothamnium alternans are developed to a relatively high degree. Very local reactions by individual zooids result in the retraction of the ciliated oral disc. More general responses evoked by altered salinity of the water, sudden contact stimuli, etc., are expressed by an extremely rapid contraction of the stalk and the zooids. Gradual relaxation or extension ensues within fifteen to thirty seconds unless the stimulus is sustained. This manifestation of irritability constitutes one of the major technical difficulties for experimental work. When the colony is contracted, it is practically impossible to distinguish a given zooid or to cut the stalk at a desired point. Fortunately for the experimental work the stalk alone may be touched gently without disturbing the colony. A series of attempts to inactivate the colonies by subjecting them to a variety of narcotics in various concentrations were quite unsuccessful. Freezing tempera- tures did not modify the contractions to any great extent.
SUMMARY
1. The frond-like colonies of Zoothamnium alternans develop ac- cording to a well-determined pattern. This development is described with particular reference to the origin of the heteromorphic zooids.
2. An asexual propagative zooid (ciliospore) detaches from a mother colony, swims away, and settles down to form a new colony. When affixed to the substrate the ciliospore begins to secrete a peduncle or
NORMAL DEVELOPMENT IN ZOOTHAMNIUM ALTERNANS 129
stalk. THe first portion of the peduncle is flexible but not contractile. From a point some 200 to 300/i above the hold-fast, the ciliospore suddenly begins to produce a contractile cord in the core of the hyaline stalk substance. Distal to this point the contractile cord is an integral part of the stalk; it is continuous from branch to branch and from cell to cell.
3. When the stalk is approximately 500/x long the ciliospore divides unequally. The larger daughter remains axial in position. It re- presents the first generation of the terminal macrozooid series (TM. No. 1). The smaller lateral cell is the initial zooid of the first branch, the median microzooid A .
4. Successive divisions of the terminal macrozooid produce each time a terminal macrozooid of the next generation and a median micro- zooid. The latter are strictly alternate in position: they alternate on right and left sides of the primary axis at successive nodes. As many as thirty-three generations of the terminal macrozooid have been observed.
5. The first division of the initial branch zooid gives rise to an axial microzooid and a lateral stem cell. The stem cell generates alternating zooids of the main branch strain whereas the axial microzooids on some of the branches represent the presumptive ciliospores.
6. At least four types of zooids comprise a colony: (1) a single terminal macrozooid at the apex of the primary axis, (2) many common microzooids scattered along each branch axis, (3) a terminal zooid at the tip of each branch, and (4) immature ciliospores which arise from axial microzooids on some of the branches. During epidemics of conjugation two more types may be formed: the terminal macrozooid may be transformed into a macrogamont, and common microzooids at random positions may metamorphose into migratory microgamonts. The spatial distribution of the heteromorphic zooids is described in some detail.
LITERATURE CITED
CLAPAREDE, E., AND J. LACHMANN, 1858-60. Etudes sur les Infusoires et les
Rhizopods. Mem. de Vlnstiiut Genevois, 5-7. FAURE-FREMIET, E., 1930. Growth and differentiation of the colonies of Zootham-
nium alternans (Clap, and Lachm.). Biol. Bull., 58: 28. FURSSENKO, A., 1929. Lebenscyclus und Morphologie von Zoothamnium arbuscula
Ehrenberg. Arch. f. Protist., 67: 376. GRAVE, B. H., 1933. Rate of growth, age at sexual maturity, and duration of life
of certain sessile organisms, at Woods Hole, Massachusetts. Biol. Bull.,
65: 375. WESENBERG-LUND, C, 1925. Contributions to the biology of Zoothamnium
geniculatum Aryton. D. Kgl. Danske. Vidensk, Selsk. Skr., nature . og.
math. Afd., 8. Raekke, X: 1.
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS
F. M. SUMMERS (From Bard College, Columbia University)
INTRODUCTION
Interest in morphogenetic studies has long centered about the part- whole relationships obtaining throughout the formative period in organic development. In so far as a system comprising several parts is concerned, the term organization implies the existence of integrating factors that condition to some extent the limits and direction of regional specialization. A remarkable number of investigations on metazoan organizers have already demonstrated the importance of extrinsic factors upon determination in specific parts. It was felt that additional information about these factors could be gained by applying operative techniques to an animal type in which, presumably, the interrelationships have not attained so great a degree of complexity.
The principal endeavor of this work was to investigate some of the qualitative and quantitative aspects of growth and differentiation in Zoolhamnium colonies, regulating after the removal of actively growing (distal) parts, and to compare regulative behavior with the normal development already described by Faurc-Fremiet (1930) and Sum- mers (1938). The results have made it possible to offer a rough map of potencies and prospective values of individual cells at various positions in the colonial pattern and to indicate some of the changes in the expression of inherent potencies which may be induced by experi- mental means.
It is a pleasure to thank Professor L. R. Cleveland of Harvard University for extending to the author the privileges of his laboratory during January, 1937.
MATERIALS AND METHODS
The materials and techniques used in this study are similar to those described in the previous paper and need not be repeated here. The only additional detail pertains to the method of shearing the stalk. Two fine scalpels made from No. 9 sewing needles were used for the purpose. One was brought to rest against the surface of the stalk and the other sheared against it in scissors fashion.
The plan of attack is by no means new, but the type of organization
130
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 131
dealt with seems to promise a fresh approach to the current problems of form determination. A colony of Zoothamnium alternans is ad- mirably adapted to work of this kind by virtue of the regularity and precision with which the characteristic colonial pattern develops. The alternating arrangement of the branches and cells makes it compara- tively easy to follow the history of any one cell throughout the course of its development for evidences of growth, division, or differentiation. The spatial relationship of the cells minimizes to a great extent some of the factors so difficult to evaluate for compact tissues. Crowding effects such as mutual contact, pressure, etc. (Peebles, 1931) are of no great concern here. Then, too, the separated cells are uniformly bathed by an almost constant medium, filtered sea water. Physio- logical relations between them are effected through a well-defined channel, the stalk with its neuro-muscular cord (Faure-Fremiet's "cordon central"), rather than through the general expanse of juxta- posed cell membranes.
REGULATIVE DEVELOPMENT
Standards of Judgment. — Colonies maintained for several days on slides are apt to be attacked by internal parasites or covered by plant growths of one kind or another, especially in the basal regions. When operations are made the axial growth of a colony is retarded for an average of 23.1 hours pending the formation of a new terminal macro- zooid. It is during this period of arrested development that adverse environmental conditions are liable to bring about an incapacitation or loss of important zooids before a decision relative to the success of the operation can be reached. A small proportion of the successful operations shown in Tables I and II do not appear in subsequent tables because they were destroyed or abandoned after indubitable signs of new terminal macrozooid differentiation had appeared but before descendants were produced. In the absence of a regenerate, the responses were recorded only when all of the structural character- istics of the new terminal macrozooid were established and, in addition, the "activated" branch developed an anterior flexure. In conse- quence of the stalk curvature the new terminal macrozooid assumes the apical position upon an anteriorly directed axial stalk. The point of curvature marks the node (Figs. 3 and 5) at which the stalk suddenly increases to a diameter approaching that of the original axis.
Regenerative responses were arbitrarily called negative only when one of the following conditions were realized: (a) there was no activity for at least 48 hours; (b) in the event that mitotic activity continued for a generation or two, a minimum of 72 hours was allowed for signs
132
F. M. SUMMERS
of regulatory activity; (c) when the terminal branch zooid in the line of succession metamorphosed into a migrating zooid of some kind. The thirty-seven negative cases shown in Tables I and II were main- tained for a mean time of 94.3 hours after the last division, with ex- tremes of 52 to 212 hours.
TABLE I
|
PARTS CUT AWAY |
||||||||||
|
TM+O |
TM.-H |
TM.+2 |
TM.+3 |
TM.«-4 |
TM.+5 |
TM.+6 |
TM+7 |
Total |
||
|
- |
1 |
1 |
||||||||
|
P |
(1) |
(2) |
(1) |
(4) |
||||||
|
O |
2(1) |
1 (1) |
(D |
(D |
3 (4) |
|||||
|
N |
(D |
(D |
||||||||
|
z 0 H |
M L |
(D |
(1) (1) |
1 (2) (2) |
||||||
|
OPERA |
K J 1 |
1 2(1) |
1 |
(1) 1 |
(O 1 |
(I) |
1 (2) 2 (1) 3 (1) |
|||
|
O |
H |
1 |
(1) |
0) |
1 (2) |
|||||
|
LJ LJ |
G F |
3 (1) 4 |
1 (1) 1 |
(1) |
(0 |
4 (4) 5 |
||||
|
E |
1 |
1 |
1 |
3 |
||||||
|
D |
3 |
2 |
5 |
|||||||
|
C |
6 |
4 |
1 |
1 |
12 |
|||||
|
B |
3 (1) |
4 (1) |
2 |
1 |
(D |
10 (3) |
||||
|
A |
9(6) |
6 (3) |
6 |
2(2) |
1 |
1 |
1 |
1 |
27 (1 1) |
|
|
Total |
^7(13) |
21 (12) |
9 (3) |
5(6) |
3 (1) |
i 0) |
1 0) |
1 |
78(37) |
The distribution of regulative responses in 115 operated colonies summarized according to the number of apical branches cut away: T.M. + 0 = only the terminal macrozooid removed; T.M. -f- 1, T.M. + 2, etc. == terminal macrozooid plus one, two, or more branches removed. The letters in the left-hand column designate the branches from which the regenerates arose. The numbers of cases in which no response occurred are shown in parentheses. The first responses of the successively operated colonies are included. No attempt has been made to indicate the number of zooids on the branches dissected away; it may be estimated by referring to Fig. 1 in the previous paper.
All colonies were abandoned when the important zooids appeared to be unhealthy.
Adjustment of Descriptive Notation.— \\\ arbitrary departure from the standard notation (see Faure-Fremiet, 1930; or Summers, 1938) seems to be advisable in view of the complications arising from the
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 133
designation of generations produced by the secondary zooid near tin- point of origin of the new terminal macrozooid. To illustrate: if the new terminal macrozooid differentiates directly from 2al, then the secondary zooid 2a2 continues to generate zooids of the branch A, viz. 2#21, 2a22; 2a211, 2a212, etc., which complicates a system already difficult to summarize briefly.
At this point it is proposed to adjust the terminology so that the new terminal macrozooid (actually 2&1) corresponds in position to the
TABLE II
|
ORIGIN OF NEW TERMINAL MACROZOOID |
||||||||||||
|
IX |
I IX1 |
Ix |
2x' |
2x* |
3x' |
4X1 |
4x2 |
5x' |
6x' |
Total |
||
|
Q |
1 |
1 |
||||||||||
|
P |
(D |
(0 |
(0 |
(D |
(4) |
|||||||
|
O |
2 |
' (D |
(2) |
(1) |
3(4) |
|||||||
|
N |
(D |
(0 |
||||||||||
|
z o |
M |
(D |
1 |
(D |
1 (2) |
|||||||
|
1- < |
L |
0) |
(0 |
(2) |
||||||||
|
cr Ul |
K |
1 |
0) |
0) |
1 (2) |
|||||||
|
Q. O |
J |
2 |
(0 |
2(1) |
||||||||
|
Li. |
1 |
1 |
1 |
1 |
(O |
3(0 |
||||||
|
o _J |
H |
1 |
(D |
(D |
1 (2) |
|||||||
|
Id |
G |
1 |
2 |
1 (2) |
0) |
(I) |
4(4) |
|||||
|
LJ |
||||||||||||
|
_l |
F |
2 |
2 |
1 |
5 |
|||||||
|
E |
1 |
1 |
1 |
3 |
||||||||
|
D |
1 |
3 |
1 |
5 |
||||||||
|
C |
8 |
3 |
1 |
12 |
||||||||
|
B |
5 (2) |
2 |
1 |
1 |
1 |
(O |
10(3) |
|||||
|
A |
1 0) |
0) |
15 (7) |
5 |
0) |
5 (1) |
1 |
27(1 1) |
||||
|
Total |
5 (1) |
2 (1) |
37 (1 1) |
19 (6) |
3 (2) |
8 (6) |
4 (2) |
(3) |
(4) |
(1) |
78(37) |
Lateral distribution of operations. This table summarizes the data from Table I according to regulation by branch generations irrespective of the amount of colony cut away. The left-hand column represents the branches from which the regenerates developed; headings indicate the position and generation of the zooids which produced the new terminal macrozooid. The negative responses are shown in parentheses.
original terminal macrozooid (T. M. No. 1), but is distinguished from it by a prime (') symbol. The lateral branch zooid 2a2 becomes A', the first zooid of a new branch at the original A level (Fig. 1).
This adjustment is valid if subsequent products are to be treated descriptively as new primary branches rather than as branches of the second order on the first branch below the cut. This is nothing more than a manipulation of terms to facilitate comparison of normal and regulating colonies.
F. M. SUMMERS
Trauma. — For a number of reasons it appears probable that traumatic shock effects are not significant factors in post-operative adjustments of colonial form. Cells adjacent to the cut areas and elsewhere soon expand and feed as before. Processes of mitosis or
REGENERATING PORTION
( TM.2' ) /
<3a!;
\_y
/ \
i TM.I' i
* /
•~^<
?2a')
FIG. 1. Diagram of regulative development. The revised notation for desig- nating cell lineage in regulating portions is indicated within the circles. The symbols in parentheses under the circles illustrate how cumbersome the conventional termi- nology would soon become if applied to the regenerates. The diagram shows the lineage of a regenerate from branch .1. It the principal axis is severed between branches A and fi at a time when there are three cells on branch .1, the regenerate usually arises from the terminal branch zooid 2al as illustrated. The sub-terminal zooid (2a2) assumes the terminal branch position A' and continues to generate branch cells. The new branch B' is produced by the first division of the new terminal macrozooid T.M. \'. The parts produced after cutting are drawn with broken lines.
differentiation, in progress at the time of operation, continue without perceptible interruption. Or these processes may begin at varying intervals after cutting, in any of the branch zooids, except that which becomes the new terminal macrozooid. Furthermore, when there is
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 135
no regulation, or when the median microzooid divides one or more times before the new terminal macrozooid is recognizable, the first branch below the cut continues to generate common zooids as before. Relative to division rates, the available data indicate that, exclusive of the one which bears the presumptive terminal macrozooid, branch growth is not perceptibly altered after cutting. As a rule there is a lag in the development of the activated branch pending the differentia- tion of a new growing point.
The only effects of mechanical disturbance are evidenced for a very short time after cutting by a state of irritability during which the contraction of decapitated colonies is frequent, irregular, and some- times tetanic. But normal overt behavior is resumed within a few moments when the stalks are shorn cleanly at internodal points. Operations were considered acceptable only when the normal reactions were regained within a relatively short time. Cases where only the neuro-muscular cord of the stalk was injured are to be treated in another section.
Distribution of Operations. — A general resume of the experimental results in terms of initial regulative responses in Zoothamnium alternans is given in Tables I and II. Of the 144 protocols at hand (acceptable operations), 78 yielded positive responses and 37 were negative; the remainder were inconclusive according to the standards chosen and are omitted in the digests.
Table I summarizes the responses to various types of cuts made at the several levels along the principal axis irrespective of the number of generations on the regulating branches. In Table II the same protocols are tabulated according to regulative responses by the various branch generations without regard to the number of branches or zooids removed. In this table the negative cases show only the zooids which were expected to reconstitute the axial growing point; some of these, failing to regulate, continued to develop laterally with- out further differentiation.
The symbols X and x are used to indicate generations on a general- ized branch (Summers, 1938). With reference to a specific branch, e.g. branch D, the symbol lx refers to the microzooid Id, and IX means axial microzooid \D.
Simple Cut-offs. — In general when the terminal macrozooid was cut off, the terminal cell of the first branch below the cut differentiated into a new, well-defined terminal macrozooid whose first and subse- quent divisions proliferated the alternating median microzooids (initial branch cells) of the regenerate. Some operations were made at a time when a single cell, the median microzooid X, represented the
136
F. M. SUMMERS
adjacent rudimentary branch. In all such instances recorded at least one division followed without perceptible delay, thus producing an axial microzooid 1A" and the presumptive terminal macrozooid l.v. More frequently the removal of a terminal macrozooid left two or three cells on the last branch (Fig. 1). Simple cut-offs leaving this branch with more than three cells were rarely possible for the reason that a division of the terminal macrozooid usually preceded the third division on the adjacent branch.
There were a number of cases where, after an operation, the ap- pointed branch continued to develop at a normal rate for one or more
A
FIG. 2. A. Colony 37/3-5 drawn approximately 53 hours after the operation. The apical cell was destroyed at the two-cell stage of development. The remaining microzooid A divided twice before the new T.M. differentiated from 2a>. Its first division produced the initial cell of the branch B' '. The alternate daughter (2a2) extended the original branch, and the axial microzooid I A persisted without further change. X 250.
B. Schematic representation of the resulting growth.
generations before another terminal macrozooid differentiated (Fig. 2). In approximately 12 per cent of the regulating colonies two divisions followed the operation, within normal time limits, before indications of a new terminal macrozooid appeared. In one case there were three pre-differentiational divisions.
Regulation from the terminal branch zooid l.v following simple decapitation of the growing point occurred in a high percentage of the cases (see Tables I and II). As far as the various levels along the primary axis were explored, the cells of the first few branch generations exhibited relatively frequent regulative responses. For tin- small number of operations made at high levels on old colonies, the transfer-
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 137
mation of l.r into a new terminal macrozooid occurred with about the same frequency and with as much dispatch as for earlier periods, i.e. lower levels on younger colonies. The metamorphosis of the acti- vated microzooid \x at levels above E required no more than the average time necessary for differentiation of apical cells produced at lower levels. This appears to correlate with the more or less uniform rate of normal axial growth (Faure-Fremiet, 1930). Likewise 2xl zooids on the various branches responded in the majority of trials. Zooids of the 3xl generation or later failed to regenerate above the mid- region of the experimental colonies.
In order to test the responses of a branch cell of the third generation it was necessary to cut away the newer branches which had formed above it along the main axis. In tests of the fourth or later branch generations, a relatively large part of the colony had to be removed.
Compound Cuts. — When the terminal macrozooid and the terminal cell of the last branch were removed, the terminal cell of the second preceding branch was frequently induced to differentiate into a new terminal macrozooid. This particular relationship obtained for a limited number of successive branches and even then was unpredicta- ble. For instance, the terminal macrozooid was sometimes produced by the sub-terminal (a secondary) microzooid of the newest branch despite the presence of a healthy terminal cell on the penultimate branch. In rare cases the latter assumed the regulative function in the presence of a complete uninjured branch between it and the cut-off.
The number of regenerates obtained from the sub-terminal branch zooids is given in Table II. Microzooids of the order 2x2 responded up to the level of branch /, whereas sub-terminal zooids of the fourth generation (4^2) did not respond at all. It is also certain that some colonies did not regenerate from either the sub-terminal zooid of the first branch below the cut or from the terminal zooid of the next ad- jacent branch. These branches continued to develop in a normal fashion for one or more generations without attempting to produce a new terminal macrozooid.
The reactions of the older segments of well-developed colonies were tested by means of extensive "cut-backs," colonies cut off at some more basal internode. The data obtained (Table II) suggest an inverse relation between the number of regulative responses and the age of the activated zooids, i.e. the frequency of responses diminishes as the number of lateral generations increases. The frequency of negative cases even in younger generations increased in the high levels, which is probably an expression of the fewer generations re- quired to bring the more distal branches to full development. The
138 F. M. SUMMERS
same data arranged according to amount of colony removed (Table I) show that basal branches may differentiate a new terminal macrozooid after as many as seven branches plus the apical zooid are cut away.
The cut-back experiments were successful only as regards the demonstration of initial reactivity to surgical alterations. The cases in Tables I and II where relatively large portions of the colonies were removed show only that the prospective values of certain zooids along branch axes may or may not be modified, depending upon the amount of colony dissected away. The capacity of the responding zooids for sustained growth is not known because all of the colonies cut back four or more branches had to be discarded before the regenerate attained full growth. Indeed, some were maintained under experi- mental conditions only long enough to produce a new apical cell. The chief difficulty is referable to the fact that the basal branches were the first to be attacked by vegetable growths propagating over the surface of the slide. The affected basal zooids were shed before the colonies reached maturity. For this reason the zooids of branch A on colonies with eight or more branches were usually unsuitable for testing. The age of the colony at the time of operation plus the additional time required for the differentiation of one of its zooids gave to the parasites an advantage that was too frequently fatal to the experiment.
Regulation from the Axial Microzooid Series. — Although nearly 9 per cent of the regenerates sprang from the axial microzooid (\X) series, their regulative behavior was capricious and could not be in- duced at will. One of the most striking facts in this connection was the origin of new growths from LY or descendants on complete, uninjured branches (Fig. 3). One originated from the third branch below the cut. I Miberate attempts to activate a given axial zooid by eliminating all other zooids on the branch resulted in (a) no further developmental activity, or (b) regeneration from some zooid on the next lower branch. A three-cell colony which was trimmed down to a single cell, the axial microzooid I A, remained without further change for 165.5 hours; its contractile and feeding responses appeared to be normal for the entire period of observation. Similarly, the corre- sponding zooids on branches E, L, M, and P of other colonies were tested without avail. In each case the regenerate developed from zooids on the next lower branch and is so recorded. On the other hand, it has been demonstrated that the axial microzooids are capable of regenerating (see IX and \Xl in Table II).
The protocols show that IX or some of its descendants carry termi- nal macrozooid potencies. In some of these cases the regenerates
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 139
formed directly from IX without leaving ciliospore-producing cells on the activated branch. The others regulated after \Xl and IX2 were formed; IX1 gave rise to the new terminal macrozooid while IX2 produced one or, after division, two typical ciliospores.
Regulation after Successive Cuts.— The successive operations were of two general classes: (a) progressive, in which the second and third
FIG. 3. A. Drawing of colony 3/J-2 approximately 72 hours after cutting. In this case the regenerate was produced by the axial microzooid on the intact branch /. Axial microzooid II divided before the new terminal macrozooid differentiated: I/1 regulated, leaving I/2 near the base of the new axis. An increase in diameter of the new axis is evident near its junction with the original colony axis. X 250.
B. Condition of the apical end of the colony before cutting at the point indicated by the arrow.
C. Schematic representation of the resulting growth at the time the drawing was made.
cuts removed regenerated parts distal to the preceding operation (Fig. 4) ; and (6) regressive, where the entire regenerate plus additional parts of the original colony were cut off. In the first group the second and third regenerates were themselves products of regenerated seg- ments. Those of the second group developed from some more basal
140
F. M. SUMMERS
cell of the original colony. In this respect they were similar to the extensive cut-back types and subject to the same technical limitations.
The regulative activity of the zooids of the second and third order regenerates in progressively cut colonies was similar in most respects to that evoked after simple cut-offs. In each of the nine cases studied the same morphological pattern and, as far as the limited number of cases permit judgment, similar developmental rates obtained after the second and third operations.
Incomplete Section of the Stalk. — In a few cases out of many trials a local injury to the neuro-muscular cord was effected without destroy- ing the continuity of the cortical hyaline stalk substance. A re-
FIG. 4. A. Development after two successive operations (54 hours after the first cut). The original axis lies on the right. The severed peduncle of the regener- ated terminal macrozooid may be seen near the base of the second regenerate. Axial microzooid 1C was badly parasitized; it dropped away soon after the drawing was made. X 250.
B. Schematic representation of A. The cuts are indicated by arrows. *Para- sitized zooid.
quisite degree of compression between the needles caused the cord to break down into a series of irregular protoplasmic droplets, some of which appeared to be independent of any attenuated membranous connectives.
Several interesting facts we're brought out by this type of opera- tion. The severed part of the neuro-muscular cord did not recover from the injury, i.e. the structural or functional cont inuity between the separated parts was not re-established. After the injury there was no subsequent degeneration of the cord in either proximal or distal parts. The functional unity of the whole was permanently impaired. The
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS
proximal and distal parts contracted independently of each other and, as esteiblished in one case, the distal portion continued to grow and differentiate in the manner of an intact colony, whereas the proximal part regenerated a new primary axis from a zooid below the injury (Fig. 5).
More substantial data are obviously required before definite conclusions can be drawn. Nevertheless these results do suggest interesting possibilities for further study. A thoroughgoing investi- gation of the contractile, transportative, and transmissive properties of the neuro-muscular cord may lead to a further elucidation of coordi- nating factors in colony formation. At least here is a clear indication that, whatever the physico-chemical nature of the integrative factors, they are probably mediated through the substance of the cord.
^m A.
FIG. 5. A. Branch C of colony 9B4 56 hours after injury to the neuro-muscular cord (drawn from above). The original colony of six branches was pinched in the mid-region, isolating ABC from DEF plus the terminal macrozooid. The terminal cell on branch C (3cl) at the time of the operation differentiated into a new terminal macrozooid which produced two new branches as shown. Note the increased diameter of the stalk just lateral to the second zooid. This marks the position of the microzooid 3cl at the time of injury. X 250.
B. Schematic representation of branch C as drawn. Axial microzooid 1C was accidentally cut away from the position marked (x).
Several completely isolated fragments were followed for a time by transferring every few hours to fresh filtered sea water. Nothing of unusual interest occurred in their development. Growth, differ- entiation, and regulation in progress at the time of cutting continued as before for the few generations that were followed. They did not re-attach to the substrate but developed as free-swimming fragments. Their growth capacities or the minimum size necessary for survival were not investigated.
Differentiation in Regenerated Parts.— The regenerates formed on decapitated colonies, after single or successive operations, are capable of producing any of the six types of heteromorphic zooids previously described (Summers, 1938). According to the data compiled from
142
F. M. SUMMERS
77 protocols the type, number, and distribution of zooids on the regenerated parts compare favorably with the control colonies. The regenerates consisted of a new terminal macrozooid, varying numbers of common microzooids, a terminal cell at the tip of each branch, and one or more potential ciliospores, depending upon the degree of develop- ment following an operation. Macro- and microgamonts likewise differentiated on regulating parts with about the same frequency and vertical distribution as for corresponding regions of normal gamont- producing colonies. Regenerates sometimes differed from the con- trols in respect to the branch generation involved in the production of ciliospores and microgamonts. In normal colonies the ciliospores developed not earlier than the \X generation and the microgamonts only from 2xl or succeeding generations. On the regenerates one or the other of these two types of migrating zooids frequently developed from the initial branch zooid (Fig. 6B). This tendency towards
..
FIG. 6. A. Terminal macrozooid 13' (above branch M) differentiating into a ciliospore 204 hours after cutting. The colony was sectioned between branches E and F. X 250.
B. Metamorphosis of the median microzooid S' into a microgamont. I he drawing was made 127 hours after regulation from the axial microzooid 1O. X 250.
earlier differentiation obtained not only in the young regenerates on immature colonies but also in those derived from older (basal) zooids of nearly mature colonies.
Another noticeable deviation from the established norm was tin- occasional metamorphosis of the terminal macrozooid into a migra- tory zooid, either ciliospore or microgamont (Fig. 6A), thus bringing to a close the growth along that particular axis. The terminal macro- zooid of the regenerate may differentiate directly into any of the three reproductive zooids: microgamont, macrogamont, or ciliospore. Faure-Fremiet (1930) reported the formation of the latter type in run- instances during the normal development of this species but the process was unaccompanied by the endomictic process which he described for normal ciliospore development.
While studying conjugation in Zodthamnium arbuscula, Furssenko
FORM REGULATION IN ZOOTHAMN1UM ALTERNANS 143
(1929) observed that a local injury to one of the main branches affected the zooids distal to the region of injury, inducing many of the micro- zooids to metamorphose into microgamonts ("microconjugants"). Similar effects on the whole colony were induced by unfavorable environmental conditions, e.g. inanition or lack of oxygen. He did not observe a regulatory response subsequent to the injury.
Regulation after Conjugation. — From the following fragmentary account of the growth activity manifest in conjugating colonies it is at once clear that this aspect of development alone constitutes a lengthy research problem. Only ten of the colonies whose lineage was being followed happened to conjugate so that a detailed analysis
FIG. 7. A colony several days after the onset of conjugation at the level of branch N. The ex-conjugant divided into a cluster of large zooids, one of which differentiated into a new T.M. whose further development prolonged the main axis. The exact lineage of these cells is not known. X 250.
of the process is not immediately available. Much of that which follows is based upon conjugants observed among the adventitious growths on the culture slides whose histories are but imperfectly known. Conjugation is introduced here because it affords one clue to qualitatively different physiological relations between the apical zooid, the conjugant in this case, and a large area of the subordinate regions of the colony.
So far the results obtained from regulation experiments bespeak a regular functional correlation between the single cell in the apical position and the zooids in sub-adjacent regions, such that the latter
144 F. M. SUMMERS
are subservient to the former. Potentialities known to be present in zooids of a lower order are presumably held in abeyance by the apical influence. Cutting away the apical region evokes a response in some zooid in a subordinate but adjacent region. The response is a differ- entiation of another apical zooid whose relations with the whole are seemingly homologous with those of the original apical cell. There may be a time between decapitation and subsequent regulation when the apical cell influences are altogether absent, yet the interim is not sufficiently great to seriously modify the observable growth phe- nomena. It is noteworthy that subordinate branches attain about the same end-point of lateral growth in control and decapitated colonies.
As stated in the previous work (Summers, 1938), the macrogamonts were observed only in the terminal macrozooid positions 3 to 24 along the primary axis. The fusion of gamonts invariably brought axial development temporarily to a close some 12 to 13 hours after the last mitosis.
The conjugant remained quiescent for periods of about four days, then divided into two moderately large zooids. One of the two ex- conjugants assumed the form of a terminal macrozooid and resumed axial development after the four-day interruption. The fate of the sister ex-conjugant is a matter for conjecture at present; some disap- peared from the colonies between observational periods while others divided into clusters of from two to seven large ciliospore-like zooids at the base of the new axis (Fig. 7). Tin- histories of these are likewise unknown. Apparently they do not propagate additional axes while associated with the parent colony. The development of ciliospores from some of the ex-conjugants in Zoothamniiim arbuscula (Furssenko, 1929) is suggestive, however.
The point to be made relates primarily to the behavior of the colony as a whole following the conjugation process. Prior to the completion of conjugation and continuing thereafter a new growth phenomenon appeared. The first three or four branches below the presiding conjugant developed out of all proportion to the average expectations (Fig. 8). The number of branch generations was in some instances greater than twice that of corresponding branches in con- trols. Moreover, many of the common lateral or secondary zooids were activated to divide one or more times, originating second order branches which, in turn, sometimes produced tertiary branches. In this way each of the first few branches below the conjugant level grew almost as individual colonies. The greatest lateral growth effect obtained nearest the conjugant and diminished basally as a gradient. The normal tendency toward a pyramidal colony pattern was thereby
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 145
reversed in the environs of the conjugant. More precise information regarding growth intensity and capacity factors in both the ex-con- jugant strain and the subordinate branches awaits further investigation.
DISCUSSION
One of the most important consequences of the work is the demon- stration of qualitatively different physiological relations between
FIG. 8. A. Schematic representation of protocol \9oe to show the dispropor- tionate development of branches K, M, and N. The lineage of branch L could not be deciphered. One of the ex-conjugant zooids at 0 produced the apical growth illustrated in the diagram.
B. The corresponding portion of the largest of the 70 control colonies.
spatially separated cells. Under normal conditions a specific pattern unfolds. When an apical region of a colony is cut away some zooid of a lower order, one whose complete developmental possibilities are otherwise never expressed, assumes the dominant generative functions, and the characteristic pattern perseveres. So far these results are intelligible in terms of what Child (1929) calls physiological correla- tion: the relations of dominance or control and subordination between
146 F. M. SUMMERS
parts. He concludes "that dominance and subordination depend primarily on quantitative, rather than specific differences in physio- logical condition and that they represent a certain aspect of a physio- logical gradient." In Zoothamnium alternans the transformation of a terminal macrozooid into an ex-conjugant initiates an entirely different developmental phase which gives another clue to the general nature of apical control. Four days after the fusion of gamonts the normal growth relations of a varying number of branches below the conjugant K'vel are upset in a rather remarkable way. Each of the three or four adjacent branches develops out of all proportion to the normal expecta- tions. The precocious mitotic activity produces secondary and even tertiary axes on the affected parts. This unusual phenomenon does not occur when the terminal macrozooid is present or when it is absent ; it is effected by some new quality in the coordinating mechanism arising in consequence of conjugation activities in one particular cell— the apical cell.
The effects obtained after decapitation and conjugation certainly suggest that the single cell in the apical position is responsible in a large measure for quantitative and qualitative regulation of the part- whole relationship and that the control varies with respect to the local activities of the parts.
There also seems to be a coordination between the cells on different branches. A new terminal macrozooid arising from one of the branch cells exerts its influence from what was formerly considered to be a branch position. Perhaps branch-to-branch coordination also ex- plains the stable activity of the variously placed cells in the interim before a new apical cell differentiates; only one of the several possible zooids regulates. The control of a terminal branch cell over the cells on its own branch can be interpreted in a similar way. As long as the terminal cell presides over a branch strain its immediate relatives remain quiescent. If it is destroyed, however, the sister at tin- sub- terminal position assumes the functional role of the lost cell.
A rather wide variation in the degree of regional correlation is suggested by those instances where a new terminal macrozooid arises, not on the first branch as usual, but on the second or even third branch below the operated level.
The axial microzooids on every branch do not differentiate into ciliospores. Loci of metamorphosing zooids occur on about every third or fourth branch. The prospective value of a single axial microzooid (LY) is predictable at a relatively early period by a marked growth in size. The growth may be taken as a criterion of at least a partial differentiation, to be completed a good many hours later
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 147
by a further increment in volume, modification of form, appearance of motile organelles, etc. A peculiar characteristic of these large cells is that they may differentiate directly into mature ciliospores or they may divide, giving origin to two zooids of unequal size, both larger than any of the common types. The larger of the two matures first; the smaller grows to the size of its predecessor before metamorphosing or dividing again to produce two mature ciliospores in succession.
Regenerates sometimes arise from either those axial microzooids which are, to all appearances, not predestined to metamorphose, or from those already partly differentiated. In the latter a potency or potencies in the process of expression apparently can be altered or superseded by others whose "urgency" toward expression is greater. The directional change in the process is referable to stimuli arising from the altered colonial organization. This is but another bit of evidence to the effect that cellular organization is dynamic and labile at certain periods and that changes going on within the cell which lead to recognizable morphogenetic characteristics are not necessarily irreversibly determined in direction. Many cases are known among the Protozoa where extrinsic, or intrinsic factors lead to periods of reorganization, varying in considerable degree for the different groups and at different periods in the life cycle. In Zoothamnium alternans we have a case where the re-direction of morphogenetic processes can be traced to an extrinsic cause: cutting the colony in the near vicinity of the cell subsequently affected.
Several significant problems arise in connection with the variable response of the axial microzooids. Why do regenerates sometimes arise from axial microzooids (IX or descendants) when as a rule the new growths are derived from the terminal branch cells? Attempts to induce regulation from \X cells by trimming away all other cells on the branch gave no positive or predictable results. Until the question is investigated further in Zoothamnium we can only interpret the variable behavior in terms of other work. Some of the merotomy experiments on other protozoa are suggestive (Calkins, 191 la, b; Peebles, 1912; Young, 1922; Dembowska, 1926; Taylor, 1928; and others). With respect to regenerative capacity these investigators were able to demonstrate progressive physiological changes in the cellular organization during the inter-mitotic period. Fragments cut at successive intervals after fission gave an increasingly high percentage of perfect regenerates. In Uronychia (Calkins, 191 la) even an emicro- nucleated fragment regenerated when cut immediately before the onset of fission. But in nearly all of the different forms studied the regenerative tendency disappeared sometime during the division process.
148 p. M. SUMMERS
f
Another line of investigation summarized by Calkins (1934) and Summers (1935) demonstrates the cytological changes in cell organelles coincident with the division process. The resorption of old and the reappearance of new motor organelles, macronuclear reorganization, etc. suggest a brief period of cellular de-differentiation. There are probably analogous processes of alternating differentiation and de- differentiation in the history of the individual zooids in Zoothamnium. Time may be one of the important factors in individual cell behavior in relation to the balance between extrinsic coordinating influences and the aggregate of intracellular activities. That is to say, the intracellular activities of a Zoothamnium cell may lead to the "fixa- tion" of specific potencies at some critical period after cell division or, conversely, a cell may be more susceptible to the coordinating influences during or immediately after a division process. The axial microzooid, for example, divides in from 20 to 70 hours after its deriva- tion from the initial branch cell, whereas the terminal cell on the branch generally divides at intervals of about 12 hours. If a decapita- tion is made at a moment when the axial microzooid is in some phase of divisional reorganization and the terminal cell in a more stable condi- tion, the former instead of the latter may be activated or excited to prolonged generative activity. The supposition should be tested by a series of accurately timed operations above some particular branch.
An explanation of morphogenetic processes in Zoothamnium al- ternans in terms of embryonic segregation at the* time of division has already been attempted by Faure-Fremiet (1930). In order to out- line several points for discussion it is essential to review briefly his cytological analysis of normal development in this species. First, as regards the early axial divisions, the first three generations along the primary axis are unequal divisions. The inequality of the resulting daughters is reflected in the assortment of macronuclear material; each time the zooids which remain in the terminal position (T<\L 1, 2, 3) receive a larger portion of the niacronuclei than the smaller branch microzooids (A, B, C). On the supposition that the enlarged end of the macronucleus apportioned to the terminal macrozooids represents a kind of segregation of chromatin material, these three unequal divisions are described by Faure-Fremiet as qualitatively and quanti- tatively differential divisions. Beginning with the fourth division (division of TM. 3) the extremities of the dividing macronuclei in .ill later axial divisions are similar in size but in each instance a bit more of the finely striated mid-portion of the macronucleus is received by the zo<")id remaining in the terminal position. All of these later divi- sions are characterized as quantitatively differential only. With
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 149
respect to the branch generations, all divisions along branches A, B, and C are similar and almost equal. The initial zooids of subsequently produced branches (D, E, F, etc.) undergo qualitatively differential divisions: the axial microzooids (ID, IE, \F, etc.) receive a greater share of the macronuclei than their lateral sisters (Id, \e, I/, etc.) although the cytoplasm in each case is distributed equally The axial microzooids (\D, IE, IF, etc.) represent the ciliospore-producing members of the colony. They undergo marked growth in size ac- companied by a disintegration and reconstitution of the macronuclei which, although not described in detail, is characterized as an endo- mictic process. Fission in the lateral sisters (Id, \e, If, etc.) is of no further interest cytologically ; these zooids constitute the main branch strains.
From the foregoing description it follows that differential division occurs at two points in the formation of all branches above C. To illustrate (Fig. 9) : D receives less cytoplasm and macronucleus than
-''.•>
--^- *•
FIG. 9.
TM. 4, but the macronucleus in both resulting zooids is qualitatively similar. The division is therefore quantitatively differential. When the initial branch cell D divides its cytoplasm is distributed equally but the macronucleus is assorted differentially because a thickened granu- lar part goes wholly to the axial microzooid \D. The division is there- fore qualitatively differential.
According to Faure-Fremiet, "It appears clearly then that during the growth of a colony of Zoothamnium alternans the two cells resulting from a division of one initial cell are never equivalent as to their 'potentialities.'" Also, "The character of the differential divisions on the main strain seems to determine the individual's differentiation of the colony; this differentiation depends not only on the individual' size, but also upon its physiological potencies." The differential divisions also appear to determine the characteristic features of the
150 F. M. SUMMERS
median individuals and of the microzooids. The latter have a limited power of growth and multiplication. Median microzooids on branches A, B, and C remain as common nutritive zooids, whereas the corre- sponding zooids or their descendants on branches above C may undergo considerable growth and metamorphose into ciliospores.
Is it to be inferred from this analysis that quantitatively differ- ential divisions determine (restrict) the subsequent power of division in branch strains and, further, that qualitatively differential divisions effect a segregation of potencies for ciliospore formation?
Regarding the first inference, all alternating median microzooids, initial branch cells, are the lesser products of quantitatively differential divisions; they can divide only so many times, according to the number of generations normally produced on the branches of which they repre- sent the beginning. \Yhile there is no doubt about the inequality of the cytoplasmic distribution between terminal macrozooids and initial branch cells for the first ten generations or more, the inequality diminishes beyond this point until the two daughters are no longer differential as regards volume of cytoplasm. An equality may be achieved as early as the tenth and not later than the twentieth axial generation. Then what of those colonies that developed eight to fifteen generations beyond the twentieth node with similar axial- lateral relations? Another point to be made pertains to the regulative capacity of the branch zooids. Those distributed in alternate posi- tions along a branch axis seldom divide further so long as the integrity of the whole colony is preserved. When an apical piece is cut away from the colony some one of the more lateral zooids on the remaining portion is capable of assuming the principal generative functions for relatively long periods of time. This behavior is difficult to interpret on the assumption that mitotic or "growth" potentialities are condi- tioned at either or both of the first two generations on the regulating branch.
The second inference may be challenged upon the grounds that segregation of ciliospore potencies does not occur at the specified division. Faure-Fremiet adduces cytological evidence of segregation at the first three axial divisions and thereafter at the division of the initial branch zooids. The latter is the fission at which the ciliospore- forming zooids are separated from the main branch strain. In the light of newer findings, the restriction of ciliospore formation to \D, \E, IF, etc. can be questioned. Ciliospores were observed to develop from axial microzooids on the second and third branches, and also from both daughters of the fifth and tenth generations on branches I) and // respectively of control colonies. Moreover, ciliospores oc-
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 151
curred on the regenerate in nearly every case of regulation from branch zooids lateral to the supposed differential division provided, of course, that they were maintained for a sufficient length of lime. The final bit of evidence comes from the demonstrated regulative capacity of the axial microzooids (1^0 on some of the operated colonies. When activated these were able to regenerate comparatively large sections of colony upon which all classes of zooids except gamonts appeared.
The spatial relationship of cells should not be minimized as an important determining factor in an organism whose cells are so charac- teristically placed. The importance of this factor in development cannot be properly valuated until the general physico-chemical nature of the integrating mechanism and the medium through which it operates, presumably the neuro-muscular cord, are more fully under- stood. It is fairly certain, however, that it is not a specific factor, for in normal uncut colonies ciliospores occasionally develop in odd positions, and the microgamonts are apt to differentiate from the common zooids at almost any position lateral to the axial microzooids. Of related interest is the work of Buchanan (1927) on the flatworm Phagocata. The region from which a piece is taken with reference to the mouth of the intact worm is of no significance in determining the location of the mouth in the regenerate. Seyd (1935), on the other hand, reported a definite degree of regional specificity in the regenera- tion of a new mouth in Spirostomum; mouths in abnormal positions in the cut organisms degenerated and new ones formed at the correct locations.
The conjugation processes in Zoothamnium arbuscula (Furssenko, 1929) and Z. alternans do not differ in essential detail. In the former the zooids at the terminal position on each of the two primary axes (Ae and Bc) becomes the "macroconjugants." Until metagamic divisions occur, further growth along these axes is arrested. Two metagamic divisions of the conjugant result in a cluster of four large zooids, two of which (A^ and C%) metamorphose directly into macro- zooids (ciliospores). The remaining two (B2 and DI) divide again, each giving rise to another macrozooid and a stem cell. A stem cell produces an additional macrozooid and a new (ex-con jugant) axis. A single conjugant therefore produces two ex-conjugant axes and from four to six successively produced macrozooids.
The behavior of small lateral branches from the main axis at nodes basal to the conjugant in Z. arbuscula compares favorably with the precocious development of subordinate branches on a conjugating colony of Z. alternans. One or two of the small lateral branches (Seitenatschen) below the conjugant develop to the proportions of
152 F. M. SUMMERS
regular main branches. In this fashion normal colonies with seven main axes are transformed, after conjugation, into colonies with nine to eleven chief axes. Furssenko refers to the new growths as "com- pensation" branches (Ersatzzweigen). It is his belief that the enor- mous growth of the macrozooids occurs at the expense of food obtained from nearby microzooids and, similarly, the compensation branches are destined to supply the energy needs of the ex-conjugant deriva- tives, i.e. the five or six clustered macrozooids and the new ex-con- jugant axes.
The chief support of Furssenko's hypothesis that ex-conjugant generations develop at the expense of adjacent regions comes from two observations: (a) the ex-conjugants themselves are not active feeders, and (6) the macrozooids in neighboring regions either fail to mature or they divide to form common zooids. The idea presupposes a mobiliza- tion and free transport of nourishment to regions active in development.
The observations on Zoothamnium alternans are not in any way contrary to a possible role of the stalk in transportative phenomena. It is also quite likely, although yet to be proved, that nutrient materials are utilized by some zooids at the expense of adjacent ones; this may be a cause contributing to growth inhibition or differentiation in nearby cells. Nevertheless it does appear that the precocious develop- ment of subordinate branches in a conjugating colony of Z. alternans is not primarily directed toward nutritional ends. In the first place, the actively developing apices of branch strains have energy require- ments which, when taken as a whole, undoubtedly exceed the demands of the single ex-conjugant or its first few non-feeding descendants. The flux would therefore be directed away from the conjugant node. Secondly, in nearly every instance recorded the unusual development of the subordinate branches was well under way before the first post- conjugant division occurred. It is problematical whether or not the change in food requirements coincident with the transformation of a terminal macrozooid into an ex-conjugant is sufficiently great to account for the relatively far-reaching alteration of the growth pattern.
SUMMARY
1. Zoothamnium alternans is a colonial protozoan of a rather special type whose constituent cells collectively possess in some degree many of the attributes of an integrated organism. Some ot tin- integrating factors can be described in general terms from the work undertaken on form regulation.
2. When the apical cell of the primary axis is dissected away from a developing colony, a cell on some inferior branch, usually the first
FORM REGULATION IN ZOOTHAMNIUM ALTERNANS 153
below the cut, will differentiate into a new apical cell. The geo- graphical limits within which positive regulative responses occur are given in the text.
3. Development of a colony continues from the newly differen- tiated apical cell. The structural and developmental characteristics of the normal colony persevere in the regenerated portion.
4. Evidence is presented to the effect that zooids retain, for a time at least, greater developmental potentialities than are actually expressed when they comprise a part of the intact colony.
5. Under varying physiological conditions in the apical region of a colony, the coordinating influences exerted upon the mitotic activity of neighboring zooids may be inhibitory (as shown by the responses evoked after decapitation) or excitatory (when the terminal macro- zooid is transformed into an ex-conjugant).
6. In the light of observations presented, the idea of dichotomous segregation or sifting out of potencies at fission is inadequate as an explanation of localization in this species. The experimental data do not confirm Faure-Fremiet's cytological account of qualitatively differential divisions at specified division nodes on the branches.
7. There is cause to suspect that morphogenetic processes in particular zooids of Zoothamnium alternans (e.g. the presumptive ciliospores), once initiated and partly expressed in visible structure, can be conditioned or modified by cuts made in some neighboring region.
8. An hypothesis is offered to account for the origin of a regenerate from one or the other of several dissimilar cells of a branch strain. The explanation is based upon the factor of time in relation to the balance between extrinsic influences and the aggregate of intracellular metabolic activities by which potentialities are realized. The cells are thought to be more susceptible to external control during the re- organizational period of mitosis. There may be a critical time in cellular differentiation beyond which the intrinsic processes are not influenced by stimuli arising in some other part of the colony.
LITERATURE CITED
BUCHANAN, J. W., 1927. The spatial relations between developing structures.
I. The position of the mouth in regenerating pieces of Phagocata gracilis
(Leidy). Jour. Exper. Zool., 49: 69. CALKINS, G. N., 191 la. Regeneration and cell division in Uronychia. Jour. Exper.
Zool., 10: 95. CALKINS, G. N., 19116. Effects produced by cutting Paramecium cells.
Bull., 21: 36. CALKINS, G. N., 1934. Factors controlling longevity in protozoan protoplasm.
Biol. Bull, 67: 410.
154 F. M. SUMMERS
CHILD, C. M., 1929. Physiological dominance and physiological isolation in de- velopment and reconstitution. Roux' Archiv. f. entu'.-mech., 117: 21. DEMBOWSKA, \Y. S., 1926. Studies on the regeneration of Protozoa. II. Regenera- tion of the ciliary apparatus in some marine Hypotricha. Jour. Exper.
Zoo!., 43: 485. FAURE-FREMIET, E., 1930. Growth and differentiation of the colonies of Zootham-
nium alternans (Clap, and Lachm.). Biol. Bull., 58: 28. Ft" RSSENKO, A., 1929. Lebenscyclus und Morphologic von Zoothamnium arbuscula
Ehrenberg. Arch. f. Protist., 67: 376. LILLIE, F. R., 1929. Embryonic segregation and its role in the life history. Roux'
Archiv. f. entw.-mech., 118: 499. PEEBLES, F., 1912. Regeneration and regulation in Paramecium caudatum. Biol.
Bull., 23: 154. PEEBLES, F., 1931. Some growth-regulating factors in Tubularia. Physiol. Zoo/.,
4: 1. SEVD, E. L., 1935. Studies on the regulation of Spirostomum ambiguum Ehrbg.
Arch.f. Protist., 86: 454. SUMMERS, F. M., 1935. The division and reorganization of the macronuclei of
Aspidisca lynceus M tiller, Diophrys appendiculata Stein, and Stylonychia
pustulata Ehrbg. Arch.f. Protist., 85: 173. SUMMERS, F. M., 1938. Some aspects of normal development in the colonial ciliate
Zoothamnium alternans. Biol. Bull., 74: 116. TAYLOR, C. V., 1928. Protoplasmic reorganization in Uronychia uncinata, sp.
nov., during binary fission and regeneration. Physiol. Zoo/., 1:1. WEISS, PAUL, 1935. The so-called organizer and the problem of organization in
amphibian development. Physiol. Rev., 15: 639. WESENBERG-LUND, C., 1925. Contributions to the biology of Zoothamnium
geniculatum Ayrton. D. Kgl. Danske Vidensk. Selsk. Skr., natiirv. og.
math., Afd., 8. Raekke, X: 1. VOIM., I). B., 1922. A contribution to the morphology and physiology of the
genus Uronychia. Jour. Exper. Zoo/., 36: 353.
Vol. LXXIV, No. 2 April, 1938
THE
BIOLOGICAL BULLETIN
PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY
QUANTITATIVE STUDIES OF THE FACTORS GOVERNING THE RATE OF REGENERATION IN TUBULARIA
L. G. EARTH
(From the Department, of Zoology, Columbia University and the Marine Biological
Laboratory, Woods Hole, Mass.)
The purpose of these investigations is to determine the reciprocal effect of two regenerating regions in Tubularia. In order properly to design the experiment it is necessary first to work out the rates of regeneration of different parts of the stem and to determine the effect of size of stem on rate of regeneration. Following this two regions of known rates are allowed to compete with one another and the effect on rate of regeneration is measured. Since it is found that the region exhibiting the lower rate is most affected, further experiments are de- signed to measure this effect. This inhibitory effect exercised by the higher rate over the lower rate has been termed physiological dominance by Child (1929) and the term "dominance" is used in this paper to describe inhibition.
When the stem of Tubularia is sectioned the hydranth differentiates in situ from the tissue of the cut end and after 30-40 hours at 18° C. the length of the primordium can be measured by means of an ocular micrometer. At this time a constriction appears at the base of the developing hydranth separating it from the rest of the stem. The perisarc being rigid, the diameter of the hydranth can be measured and the volume calculated. The time for regeneration is measured at two stages of development. t\ is the time in hours from the time at which the stem is cut to the time at which the constriction appears between the primordium and the rest of the stem. /2 is the time for complete regeneration when the hydranth is pushed out of the opening of the perisarc. In some experiments where short pieces of stem are used it is necessary to use t\ as a measure of time since the hydranth, although completely differentiated, does not emerge from the end. In most cases, however, /2 is used for the calculation of rate of regeneration.
Rate of regeneration can then be defined as the volume of the hydranth in cubic micra divided by the time in hours. R : • Trr2L/t.
155
156
L. G. EARTH
Naturally when t\ is used the rate is somewhat higher than when /2 is the measure of time. The values obtained are in regeneration units which are termed R. V . in this paper.
It is convenient to assign symbols to the ends and this has been done in Fig. 1 which shows the method of naming the hydranths. DI, D-t, Ds, etc. are used to designate the distal or oral hydranth of each piece of the stem and are numbered consecutively from distal to
|
Ill |
n\r |
Tin |
p\r |
|
|
III |
||||
|
MM |
||||
|
Mil, |
pr i |
|||
|
-TT7T |
D' |
|||
|
r^7 |
||||
|
Dr |
i |
2 |
E |
B
•^Tinr
Dr
Dr
rrimnvr
P^
PI DI
p;
Ff
P
— < thread ligature
D!
/
P!
HID!
22
i
Mil
hydranth regenerating
FIG. 1. A shows how the stems are prepared; B, C, D how halves, thirds and sixths are prepared and designated; E shows how inherent rates are determined In- tying off a small part of the stem; finally, /«", G, II show how the iL-ngth of stem is varied while the level remains constant. l)\r,D'£ • • • relative rate of distal hydranths; P\T, P'i • relative rate of proximal hydranths; D\', D-2' • • • inherent rate of distal hydranths; P\\ Pj • inherent rate of proximal hydranths. D22 and P22 are ex- amples of absolute rates.
RATE OF REGENERATION IN TUBULARIA 157
proximal levels. D\ is always the most distal hydranth of the stem. PI, Pi, Ps, etc. are used for the proximal or aboral hydranths in the same order. A glance at Fig. 1, A, B, C, will make this clear.
The stems are kept in a rectangular dish which is cooled by a glass coil through which running sea water is circulated. The stems rest on a strip of cheesecloth wyhich is stretched over a glass frame and are covered by one centimeter of sea water. Loose-fitting glass covers are placed over the top of the dish. A glass rod, which is drawn out and bent in the form of a hook, is used for picking up and transferring long stems. For short stems a pipette bent at a 45° angle is satisfactory.
Since the stems of Tubularia are very variable in the rate at which they regenerate, controls were prepared for every experiment, and experiments involving the use of different stems are never compared. That is to say, one may not compare the rates of 15-mm. stems from different experiments for they may be vastly different, as measurements show (Table II, Experiments 5, 6 and 7).
The general procedure was to select from several colonies those colonies with straight stems and no side branches. The stems were cut off at the base and placed in a large dish of sea water and further selected for similarity in length, diameter and appearance. Some stems were translucent, while others were opaque, and these two differed in rate of regeneration. After this selection, the hydranths were cut off, and the stems of the same length were selected at random for experimentals and controls. In long stems, the additional precau- tion of cutting off 3 mm. of the stem along with the hydranth was ob- served, as the region adjacent to the hydranth often regenerates at a low rate, due possibly to the use of this region in the formation of the very large gonophores.
The Inherent Rates of Regeneration at Various Regions of the Stem
The object of these experiments is to isolate various levels of the stem so that the inherent rate of regeneration can be determined. Isolation of a region from the rest of the stem can be obtained by means of a thread ligature which is tied about the stem shutting off circulation and cutting through the tissue. This technique was used by Morgan (1902) and lately by Peebles (1931). The perisarc, which is tough, does not crack but the coenosarc is completely severed so that there are no cellular connections across the ligature. If this liga- ture be applied about 2-3 mm. from the end of the stem a small piece of tissue 2-3 mm. in length is isolated and will form a hydranth. this means the rate of regeneration of a small piece of tissue at any level
158
L. G. P.ARTH
of the stem can be measured without being influenced by the stem as a whole. The situation is analogous to self-differentiation of an explant from an embryo. It might be thought that the same result could be brought about more simply by cutting off a piece 2-3 mm. in length. However, in this case one meets with the difficulty that two centers of regeneration arise at the two cut ends and bipolar forms may arise. It is impossible to get any measure of the rate at which these forms develop.
TABLE I
Inherent rates of regeneration of distal and proximal hydranths at various levels of the stem of Tubularia. Stems 30 mm. long, cut into 6 pieces and each piece ligatured in middle. L = length in n, d = diameter of stem in /*, /i = time in hours from cutting to the constriction of the regenerating hydranth.
|
L |
d |
h |
L /i |
=?•- |
|
|
/?, |
1136 |
504 |
31.6 |
36.0 |
71.6 |
|
/?, |
1044 |
488 |
34.4 |
30.3 |
564 |
|
D3 |
928 |
456 |
36.4 |
26.2 |
43.3 |
|
D^ |
880 |
432 |
39.6 |
22.2 |
33 1 |
|
D6 |
592 |
388 |
39.8 |
14.8 |
17.6 |
|
Ds |
404 |
364 |
49.1 |
8.2 |
8.7 |
|
P, |
924 |
504 |
39.1 |
23.6 |
47.0 |
|
P, |
824 |
456 |
39.3 |
21.0 |
33.9 |
|
P3 |
844 |
432 |
40.9 |
20.6 |
30.3 |
|
P^ |
704 |
408 |
47.5 |
14.8 |
194 |
|
Ps |
416 |
372 |
44.5 |
9.2 |
96 |
|
P6 |
244 |
340 |
52.8 |
4.6 |
4.2 |
The method of determining the inherent rates of regeneration at various levels of the stem is shown in Fig. 1, E, where the superscript "i" is used to designate inherent rate. Stems 30 mm. in length are first cut into 6 pieces and then each 5-mm. piece is ligatured in the middle. Thus in each 5-mm. stem the distal half is completely iso- lated from the proximal half. Since the stem which regenerates is so short, 2.5 mm., the hydranths have difficult}' in emerging and so the time recorded is t\. This is the time from cutting to the formation of the primordium of the hydranth.
In Table I and Fig. 2 the data are recorded and plot N •<!. The rates
are calculated in two ways. R\ - and also R« -• wr-L/'d where
t\
- length of primordium in micra, r • - radius of stem in micra and l\ - - time in hours. From the table it is seen that there is a gradient in the size of the hydranth and the time for regeneration. The rate
RATE OF REGENERATION IN TUBULARIA
159
of regeneration R2 of the distal hydranths D\ - - D6 thus falls off sharply from ZV ; : 71.6 R. U. at the distal end to ZV = 8.7 R. IT. at 5 mm. The lowest rate at 0 mm. is P6{ = 4.2 R. U. Thus the most distal level of the stem regenerates at almost 18 times the rate of the most
40t80
30
20
10
25
20 15
LEVEL MM.
10
FIG. 2. Rates of regeneration at different levels of the stem of Tubularia.
RI = — ; R2 = 7rr2L//i = n3- 105/hours. Stems 30 mm. in length cut into 6 pieces and
t\
ligatured in middle. Each point is the average of 10 stems, r -= radius of cross- section of stem; L = length of primordium; /i = time in hours between cutting and the formation of the primordium. The lowest rate is rate of P6- Zero mm. is at the proximal end of the stem, 30 mm. at the distal end.
proximal level. This difference in rate must be due to causes inherent in the tissue at these levels since external influences have been removed.
160
L. G. EARTH
The rate of regeneration RI has also been calculated using length as a measure of the size of the primordium in place of volume. These calculations are included here to show (hat the difference in rate of the regenerating hydranths at various levels of the stern is not due merely to a difference in the diameter of the stem but also to a difference in
70'
60-
50'
40-
UJ
r- 30-
o:
20'
10.
D
50
100
150
200
250
30O
VOLUME
FIG. 3. The effect of volume of stem upon rate of regeneration of hydrant h of Tubnlaria. D = distal hydranth; P == proximal. Rate = n3/hrs. • 105. Volume = M3- 107. Time in this experiment is 12.
the length of the primordium. What RI really represents is the rate of regeneration 7?2 divided by the cross-sectional area of the stem and therefore it corrects for the variation in diameter caused by the growth form of the stem. For obviously it I he stem is always wider at the distal end as compared with the proximal end, the apparent rate of
RATE OF REGENERATION IN TUBULARIA
161
regeneration using volume of stem as a measure of size would always be greater at the distal end even if the time for regeneration and the length of the primordium were the same as those of the proximal end. Such a difference in rate seems to me to be entirely due to morphological reasons and would not represent at all the inherent regenerative activity of the tissues themselves.
From the results (Table I, Fig. 2), however, it is clear that the difference in rate at various levels is a real one and that both the time for regeneration and the length of the regenerate vary in different regions of the stem. When RI is used as a measure of rate, the differ- ence in distal and proximal levels of the stem is not so great as R2
TABLE II
Relative rates of regeneration in Tubularia. Both hydranths are allowed to regenerate. Dr = relative rate of distal hydranth, Pr = relative rate of proximal hydranth. Rate = M3/hours- 105.
|
Experiment |
Length |
No. |
Dr |
Pr |
Thickness in n |
Vol. Stem |
|
|
D |
p |
||||||
|
mm. |
M3. 10' |
||||||
|
1 |
5 |
6 |
68.5 |
0 |
— |
— |
— |
|
2 |
5 |
21 |
60.5 |
0 |
— |
— |
— |
|
3 |
6 |
25 |
95.0 |
4.65 |
568 |
528 |
141 |
|
4 |
10 |
20 |
55.0 |
7.3 |
488 |
480 |
184 |
|
5 |
15 |
10 |
40.6 |
1.9 |
384 |
296 |
136 |
|
6 |
15 |
10 |
46.7 |
18.8 |
424 |
368 |
184 |
|
7 |
15 |
10 |
67.0 |
12.6 |
476 |
384 |
216 |
|
8 |
15 |
10 |
70.4 |
0 |
560 |
376 |
276 |
|
9 |
14 |
20 |
93.0 |
4.9 |
576 |
460 |
296 |
|
10 |
20 |
10 |
49.5 |
19.0 |
416 |
344 |
216 |
|
11 |
25 |
10 |
72.5 |
27.9 |
460 |
380 |
346 |
|
12 |
25 |
10 |
67.8 |
20.4 |
508 |
372 |
380 |
|
13 |
25 |
10 |
105.5 |
29.0 |
564 |
424 |
470 |
shows. The rate of regeneration of the most distal hydranths ZV is 36.0 R. U. while that of the most proximal hydranth P6 ' is 4.6 R. U., or about an 8-fold difference.
From Table I it will be noted that for pieces of the same length the volume of the stem varies at different levels because the stem is nar- rower at the base. It might be argued that the reduction in volume of the stem was responsible for the reduction in rate of regeneration. The volume of the most distal 5-mm. piece is 39.4-107/i3, while the most proximal 5-mm. piece has a volume of only 26.0-107Ai3. This is a small variation in volume compared to the change in rate from 71.6 R. U. to 4.2 R. U.
162
L. G. EARTH
The Effect of Length of Stem on Rate of Regeneration of the Ends Figure 3 shows the effect of varying the volume of the stem but keeping the level of the cut constant; i.e. using the same region of the stem but with more stem attached to it. This is accomplished by a series of ligatures as in Fig. 1, F, G, II. Inspection of Fig. 3 tells us that the rate of regeneration at a particular level is increased by in- creasing the volume of the stem but that the increase is small. For example, increasing the volume from 20.8 to 155 units produces a change in rate from 22 to 26 regenerative units for the proximal hydranths. Further increase of volume to 312 units increases the rate to 34 regenerative units. The distal hydranth behaves similarly.
TABLE III
Comparison of relative and absolute rates of regeneration. Dr = relative rate
pa _ PT
in M3/hr- 105; Da = absolute rate. Same for proximal. Dominance = - - -100.
|
Ex- peri- ment |
No. Steins |
Length |
Dr |
Pr |
Lfl |
1M |
Dominance |
|
mm. |
per cent |
||||||
|
A |
6 whole stems |
5 |
45.5 |
0 |
— |
36.8 |
100 |
|
B |
14 distal halves |
7.5 |
40.5 |
0 |
38.6 |
31.9 |
100 |
|
C D |
14 proximal halves 10 distal thirds |
7.5 10 |
23.4 52.7 |
3.9 10.7 |
24.9 57.0 |
15.1 35.9 |
74 70 |
|
E |
10 middle thirds |
10 |
36.1 |
4.4 |
26.0 |
13.0 |
66 |
|
F C |
10 proximal thirds 10 distal halves |
10 13 |
19.1 86.7 |
0 41.7 |
27.0 90.5 |
8.6 53.0 |
100 22 |
|
If I |
10 proximal halves 10 whole stems |
13 15 |
41.4 67.0 |
17.9 12.6 |
53.6 63.0 |
28.5 26.8 |
38 54 |
|
J |
10 whole stems |
15 |
46.7 |
18.8 |
40.4 |
20.4 |
7 |
|
K |
10 whole stems |
20 |
49.5 |
19.0 |
51.2 |
28.9 |
34 |
|
L |
10 whole stems |
25 |
105.5 |
29.0 |
108.0 |
47.4 |
39 |
|
M |
10 whole stems |
25 |
72.5 |
27.9 |
66.8 |
34.0 |
20 |
Compare these small changes in rate with the increase obtained with change in level (Fig. 2). It is of interest here that the total increase in rate for the distal and proximal hydranth, when the 312 units of stem are used, is the same and amounts to about 12 units. This value may express the amount which the middle of the stem contributes to the ends in regeneration.
The Relative Rates of Regeneration of Pieces of Tubular ia Knowing the differences in inherent rates of regeneration and also the effect of volume on rate, we are in a position to study relative rates and the factors that determine these rates. The rate is termed , "relative rate" whenever both ends of a piece of stem of Tubularia
RATE OF REGENERATION IN TUBULARIA 163
are allowed to regenerate without a ligature. (See Fig. 1 ; A, B, C, D.) The superscript ' V is used. In this case two hydranths are regenerat- ing and are doing so in relation to one another, since they are connected by the stem. Here competition and resultant dominance come in, especially with short stems.
Relative Rates of Whole Stems. — In Table II the relative rates of whole stems of different lengths have been recorded. The stems are prepared as in Fig. I, A. The dominance of the distal regenerating end becomes greater as the stem is cut shorter, so that at 5-10 mm. proximal hydranths often do not appear. As the length of the stem is increased, t-he rate of regeneration of the proximal hydranth increases. Other factors, such as thickness and volume of stem, play a part.
Relative Rates of Regeneration in Distal and Proximal Halves. — When the stem of Tubularia is cut into halves, four hydranths may develop, which we designate as D\, D2, PI and P2, as in Fig. 1, B. The rates for such stems of various lengths are given in Table III. To show the steepness of the gradient and also the phenomenon of dominance, the relative rates are plotted in Fig. 4. In long stems (26 mm.), there is very little difference in the relative rate of regeneration of the distal hydranth of the proximal half, D2, and the proximal hydranth of the distal half, PI. This is to be expected if there is no dominance, since these are adjacent levels of the stem (Fig. 1, B). The rapidly develop- ing distal hydranth, D\, in the distal half does not inhibit the proximal hydranth, PI, any more than the slowly developing proximal hydranth, PZ, in the proximal half inhibits the distal hydranth, D2. There is thus no apparent dominance exerted in the ordinary sense, i.e. as an inhibition exercised by a more rapidly regenerating region.
However, in shorter stems dominance appears, and the relative rates of regeneration of the distal hydranth of the proximal half, Dz, and the proximal hydranth of the distal half, PI, are then quite different. In Fig. 4 the rate of D2 ---- 23 .4 R. U., PI - 0. In the distal half the distal hydranth D\ is completely dominant over the proximal PI. In the proximal half of the stem the distal hydranth, Z>2, is only partially dominant over the proximal hydranth, P2. The rate of P2 = 4.0 R. U.
Finally, in Fig. 4, we have plotted the rates of regeneration of the six possible hydranths, Dlt Z>2, D3 and PI, P2, P3 of Fig. 1, C for 10-mm. lengths of stems originally 30 mm. long. The graded differences in rate of regeneration in different regions of the stem are shown by the rates of the distal hydranths and also by the proximal hydranths in Table III, Experiments D, E, F. It will be noted, however, that at any level at which stem is cut the distal hydranth of that level re- generates much more rapidly than a proximal hydranth at the same
164
L. G. BAKTH
level. This means that the distal regenerating hydranths in each third of the stem are inhibiting the proximals in that same third. The situation can be summarized by saying that D{ > Ds > D3r and /Y • P,r • /Y, but that Ds • Pir and D3r > Ps, showing that both D\ and D» are dominant. JY is 0, showing the dominance of D3.
90-1
80-
10 15
LENGTH MM.
20
25
30
Fu;. 4. Relative rates of regeneration of cut steins. Squares = stems 26 nun. long cut into halves; triangles = stems 15 mm. long cut into halves; circles = stems 30 mm. long cut into thirds. Df, D-f, P\r, etc. = relative rates of respective li\- clranths. Rate = M3/hrs-105. The dotted line serves to connect adjacent regions of a cut stem.
RATE OF REGENERATION IN TUBULARIA 165
The plot of rates in three sets of stems in Fig. 4 shows quite clearly that in each case the gradient is steeper in the upper (distal) levels of the stem. The slopes of the gradient in the upper levels of the three sets of stems are approximately parallel, which seems to indicate that the drop in rate per unit length of stem is about the same in different stems. It should be pointed out that the gradient here will be steeper than the gradient in inherent rates (Fig. 1), since dominance lowers the rate of proximal hydranths.
The Reciprocal Influence of Two Regenerating Regions
In stems of Tubularia from 10 to 30 mm. in length, a hydranth usually regenerates at both ends, and it is the purpose of the following experiments to show how the rate of regeneration changes when either the distal or proximal hydranth is prevented from regenerating by means of a ligature allowing the hydranth at the opposite end to re- generate free from dominance. These rates have been termed absolute rates, Da, to distinguish them from relative rates, Dr. When only one end of a stem is allowed to regenerate, the rate is termed "absolute rate" in contrast to relative rate. This absolute rate of one end can be determined by simply tying off the opposite end of a stem, which prevents regeneration at this end. This rate tends to be the maximal rate of regeneration of a given region since the region has the entire length of the stem affecting its rate and is also independent of a second regenerating region (Fig. 1, //, D-). It will also be seen that the abso- lute rate is simply the inherent rate plus the increase due to the addition of the stem.
The Influence of the Distal Regenerating Region on the Proximal End. -In general (Table III), the shorter the stem the greater the inhibiting effect of the distal regenerating region, and the following cases are arranged according to the length of the stem. In each experiment three lots of stems were used. The first lot regenerated without a ligature, giving relative rates Dr and Pr. The second lot regenerated with the distal end tied off, giving the absolute rate of the proximal end Pa, while in the third group the proximal end was ligatured, giving the absolute rate of the distal end, Da.
Without exception (Table III), the absolute rate of any proximal hydranth is higher than its relative rate. Indeed, in some cases the relative rate may be 0, while the absolute rate is fairly high, 36.8 units (Experiment A}. This method of treating the proximal end furnishes a measure of dominance, for if \ve know the absolute rate and the relative rate we can calculate the percentage reduction due to the distal end, i.e. the inhibition exercised by a dominant region. In
166
L. G. EARTH
Table III, last column, this percentage has been calculated, and dominance varies from 100 per cent in short pieces to 7 per cent in longer pieces.
The Influence of the Proximal Regenerating Region on the Distal End. —In contrast to the proximal hydranth, the distal hydranth is affected little if at all by the ligature at the proximal end. This is true in the upper levels of the stem. Thus, in Table III, Experiment B, the rela- tive rate of the distal hydranth, Df — 40.5 R. U., while the absolute rate of the same hydranth, D\a --- 38.6 R. U. and also, Experiment L, Dir •• 105.5, Dia - 108.0. However, in lower levels of the stem, the absolute rate of the distal hydranth is usually greater than the relative rate. Experiment F, the relative rate of the distal hydranth of the
TABLE IV
The relation of the inherent rates of regeneration to the relative rates in pieces
M3 of Tubularia of varying length. Rate = — • 10s. D' = inherent rate of distal
hydranth; P1 = inherent rate of proximal hydranth; Dr = relative rate of distal hydranth; Pr = relative rate of proximal hydranth.
|
Liga- |
Liga- |
No 1 i«a |
No I.1LM |
Ratios |
|||
|
r>i |
PI |
ture |
ture |
||||
|
Dr |
Pr |
DlIPl |
Dr\PT |
||||
|
mm. |
|||||||
|
25 |
54.6 |
22.0 |
72.5 |
27.9 |
2.46 |
2.60 |
Chiefly partition, no dominance. |
|
20 |
38.2 |
13.9 |
49.0 |
19.0 |
2.74 |
2.60 |
Chiefly partition, no dominance. |
|
14 |
81.5 |
22.6 |
93.0 |
4.9 |
3.61 |
19.0 |
Dominance incomplete. |
|
5* |
71.6 |
47.0 |
78.0 |
0 |
1.53 |
00 |
Dominance complete. |
* /i used in place of /2-
proximal third of the stem Dsr • 19.1, the absolute rate of the same hydranth, D3a ---- 27.0; Experiment II, Df •• 41.4, D2a ---- 53.6. This is extremely interesting, because it shows that the proximal hydranth does have an inhibiting influence over the distal hydranth in lower levels of the stem. In proximal regions of the stem thru there is a reciprocal influence of the two regenerating ends on each other, each tending to inhibit the other.
The Competition between Tivo Regenerating Regions Having Different
Rates
In the second section of this paper it has been shown that adding parts of the stem to either the proximal or distal end increases the rate of regeneration. Therefore the stem as a whole contributes materials to the regenerating ends, and the way in which I his material is par-
RATE OF REGENERATION IN TUBULARIA
167
titioned can be studied in stems of different: lengths which provide different amounts of the material. The following experiments were designed to determine how the distal and proximal hydranths would divide or partition the effect of adding a definite volume of the stem to the system. For this purpose the inherent rates of regeneration of the proximal and distal hydranth were found by tying 3-mm. ligatures at the ends of the stems as in Fig. 1, F and H, D\l and PI\ In the same stems the opposite ends Z>i22 and Pi22 give the rates of regeneration with 19 mm. of stem added to each end. Finally, as in Fig. 1, A, the two ends were allowed to regenerate in competition for the intervening 19 mm. of stem.
It is seen from Table IV that in long stems (20-25 mm. lengths) the addition of the middle of the stem to the distal and proximal ends
TABLE V
Increase in rate of regeneration of distal and proximal ends when the same volume of stem is added to each.
|
Total Length |
Stem 3 mm. Inherent Rate |
Volume Added M3. 107 |
Stem 22 mm. Absolute Rate |
Increase in Rate |
|
|
mm. |
|||||
|
25 |
Distal |
54.6 |
280 |
66.8 |
12.2 |
|
Proximal |
22.0 |
266 |
34.0 |
12.0 |
|
|
Stem 17 mm. |
|||||
|
Stem 3 mm. Inherent Rate |
Absolute Rate |
||||
|
20 |
Distal |
38.2 |
163 |
51.2 |
13.0 |
|
Proximal |
13.9 |
176 |
28.9 |
15.0 |
increases the rate of regeneration of these ends in direct proportion to their inherent rates. That is, the relative rates are directly propor- tional to the inherent rates Dr/Pr = D{\P\ Therefore there is simply a partition of materials without any dominance. This result would be expected if no factor other than available foodstuffs controlled the rates of regeneration.
However, in short stems (14 and 5 mm. lengths) there is no longer a partitioning of substances in proportion to the inherent rates of the distal and proximal hydranths, but rather a dominant effect of the distal end, so that it takes more than its share. Dr/Pr > D^P*. As a matter of fact, in all non-ligatured 5-mm. pieces examined in this experiment and many others, no regeneration occurred at the proximal end at all, in spite of the fact that the proximal end had an inherent rate almost as great as the distal Di/Pi - •• 1.53. This situation repre- sents complete dominance in which all of the material goes to the distal hydranth.
168
L. G. EARTH
A Comparison of the Effect of the Distal Half and the Proximal Half on the Inherent Rate of Regeneration of the Ends
It has been shown that the middle regions of the stem contribute materials for the regeneration of the two ends. The amount which the stem contributes can be measured by the increase in the rate of regeneration of the hydranths under the assumption that rate is pro- portional to amount of substances available. Thus, in Table Y, the inherent rate of regeneration of the distal end is 54.6 R. U., and when we add 19 mm. of stem, the rate increases to 66.8 R. U. or an increment of 12.2 R. U. Similarly, Pl increases from 22.0 to 34.0 R. U., or an increase of 12.0 R. U. The increase is the same for both proximal and distal ends. (See 20-mm. stems also.) It is clear that the materials of the stem can be used by either the proximal or distal end.
Now if we ligature the stem in the middle, the effect of the materials in distal and proximal halves on the regeneration of the ends may be
TABLE VI
Comparison of the effect of distal and proximal halves of a stem on the rate of regeneration of the ends. Df is the inherent rate of regeneration of the distal end. Under distal half is the rate of the distal end, with middle ligature. Similarly P{ is the inherent rate of the proximal end. Proximal half = rate of proximal end with middle ligature.
|
D* |
Distal Half |
Increase |
P» |
Proximal Half |
Increase |
|
|
25 mm. |
54.6 |
64.6 |
10 |
22 |
26 |
4 |
|
14 mm |
820 |
98 0 |
16 |
22.6 |
24.2 |
1.6 |
studied (Table VI). It is found in the two experiments available that the increase in rate is much greater in the distal as compared \\iili the proximal half. This may be taken to mean that there are more materials available in ihc distal half than the proximal. The curves for the effect of volume on rate (Fig. 3) also indicate that the distal half is more effective in increasing rate than the proximal. As volume is added to the distal end, the rate goes up sharply and then falls off. However, as volume is added to the proximal end, the rate goes up slowly at first and then sharply. The evidence seems rather conclusive that the two regions differ in their effects on rate of regeneration.
Isolation of Distal and Proximal Regions by Means of a Middle Ligature and its Effect on the Relative Rate of Regeneration of the Ends
This method of studying dominance was used in the first experi- ments, and it is complicated by the fact that not only is isolation pro-
RATE OF REGENERATION IN TUHULARIA
duced by the ligature but the volume of the stem is also reduced. The effect on the proximal hydranth is a dual one. Isolation increases the rate of regeneration, while reduction in volume decreases the rate, and the effect is a summation of the positive and negative action. Since, as we have shown, the reduction of the volume of the stem adjacent to the proximal hydranth decreases the rate to a small extent only, the chief effect is to remove the dominant region from the sphere of action, and the net result is an increase in the rate of the proximal hydranth (Table VII).
Length is an important factor, as in 25-mm. stems the change in rate with ligature is slight, while in 15-mm. stems it may be much greater, depending on the condition of the stems. This is in keeping with the fact that little dominance is exerted in long stems.
TABLE VII
Rate of regeneration of distal and proximal hydranths of Tubularia with and without a middle ligature. Dr and Pr = relative rate; Dl and Pl -- rate under conditions of middle ligature.
|
Experiment |
Length |
No. |
Dr |
Dl |
Pr |
Pi |
|
mm. |
||||||
|
1 |
25 |
10 |
67.8 |
70.5 |
20.4 |
25.4 |
|
2 |
25 |
10 |
72.5 |
64.6 |
27.9 |
26.0 |
|
3 |
15 |
10 |
46.7 |
43.2 |
18.8 |
21.0 |
|
4 |
15 |
10 |
63.0 |
64.0 |
17.5 |
23.0 |
|
5 |
15 |
10 |
40.6 |
36.4 |
1.9 |
15.0 |
|
6 |
15 |
10 |
47.6 |
38.6 |
0 |
15.1 |
|
7 |
14 |
20 |
93.0 |
98.0 |
4.9 |
24.2 |
The Mechanism of the Dominance Exerted by the Distal Regenerating Hydranth over the Proximal End of the Cut Stem
In all previous experiments a ligature was used to block the dom- inance exerted by the distal regenerating end over the proximal end. With ligature of the stem the proximal end is allowed to regenerate independently of the distal and its rate of regeneration is greatly in- creased. This ligature, however, not only stops circulation between the distal and proximal end but also severs cellular connections. Thus it is not clear whether the factor responsible for dominance is some- thing present in the circulation or something transmitted or trans- ported through the cells. Therefore it is- necessary to determine the effect on the proximal end of stopping circulation from the distal end but leaving cellular* connections intact.
Use of a Loose Ligature for Blocking Circulation. — It is not easy to
1/0 L. G. EARTH
shut off circulation between parts of the stem of Tubidaria as the cells will rearrange themselves after compression of the stem so that the circulation breaks through once more. However, there are indications which can be seen from Table Mil where the effects of a loose ligature which was tied so as to just stop circulation is compared with a tight ligature cutting through the coenosarc and breaking all cellular con- nections. The control with no ligature shows that we are dealing with stems in which the distal end is almost completely dominant over the proximal end: i.e. distal end, 38.7 R. U.; proximal end, 2.9 R. U. A tight ligature completely isolating the two halves sends the rate of regeneration of the proximal end up to 23.0 R. U. or an 8-fold increase. The distal end shows a slight reduction to 34.0 R. U. as it is cut off from the proximal half of the stem. Now with a loose ligature where cellular connections are still intact the proximal end shows an increased rate of 23.0 R. U. over controls in spite of the fact that at the end of the experiment circulation was reestablished in a few cases through
TABLE VIII
Rate of regeneration of distal and proximal ends of Tubidaria under conditions of ligature in the middle of the stem. Rate == L/t* where L is length of primordium in micra and t% is time in hours from cutting of stem to emergence of hydranth.
No. and Length No Ligature Tight Ligature Loose Ligature
of Stems DP DP DP
10 15 mm. 38.7 2.9 34.0 _'.v() 36.4 23.0
the ligature. This sort of experiment, while it appears conclusive, is not entirely satisfactory as it is difficult to control the tying of a ligature so as to cut off circulation without breaking cellular connec- tion between the two halves.
Injection of Oil to Block Circulation. — The stem of Tubidaria is about 0.5 mm. in diameter and it is relatively easy to insert a micro- pipette for injection. It is necessary merely to crack the rigid perisarc first with two pairs of sharp watchmaker's forceps after which a pipette can be inserted while observing under a binocular microscope. A small drop of paraffin oil (Nujol) is injected and after the pipette is withdrawn the rigid perisarc snaps back into place. The perisarc must not be re- moved because then regeneration will take place at the exposed surface. Controls for this type of experiment were- stems in which the perisarc was ruptured and the pipette inserted without injection. Some of the experiments where oil was injected also served as controls since the drop was sometimes too small to block circulation.
RATE OF REGENERATION IN TUBULARIA
171
Table IX records results. There are 25 stems in each sample and of the 25 controls only 2 proximal hydranths developed, making the rate 4.65 R. U. and thus showing that the distal end exerted rather complete dominance. When circulation is blocked, however, the 25 injected stems regenerate 12 proximal hydranths, bringing the average rate to 22.8 R. U., or a 5-fold increase in rate. The distal hydranth in the injected group shows a small decrease in rate, as might be expected from previous results on the use of ligatures. A ligature in a short stem increases the rate of the inhibited proximal end but decreases the rate of the dominant distal end by shortening the stem.
Only 12 out of 25 possible hydranths appear at the proximal end of injected stems and it is interesting to examine the 13 stems which did not form a hydranth proximally. Of these 13 negative cases 9 showed that the oil drop had moved from its original position at the middle of
TABLE IX
Injection of an oil drop into the gastrovascular cavity of Tubularia. Stems 6 mm. long. Twenty-five stems used in each sample. R = rate of regeneration = nT2L//2 where r = radius of stem in micra; L = length of primordium in micra and /o = time in hours required for emergence of the newly regenerated hydranth.
|
Oil Injected |
No Injection |
|||
|
Distal |
Proximal |
Distal |
Proximal |
|
|
L |
1384 282 40.0 86.1 |
528 264 50.5 22.8 |
1536 284 40.9 95.0 |
104 280 55.0 4.65 |
|
r |
||||
|
/„ |
||||
|
R |
||||
the stem into the distal end. In the 12 stems which form a proximal hydranth all but 2 showed the oil in the original position. It is clear that because the size of the oil drop varies the smaller drops do not completely block circulation and as a result they are carried by the circulation to the distal end. When the drop is larger it is held firmly in place by the endodermal lining of the gastrovascular cavity and in these cases proximal hydranths appear associated with a complete block to circulation.
In other experiments where dominance is not so complete there is a quantitative effect on rate of regeneration of the proximal end, with injection of oil into the gastrovascular cavity. It must be remembered that this effect is not so great as would be expected since not all of the oil drops are large enough to block circulation. An example is shown in Table X, where the proximal ends of control stems regenerate at a
172
L. G. EARTH
TABLE X
The effect of blocking circulation in Tubularia with oil drops. Oil injected into distal region. Controls consist in injury similar to that of injection. Twenty-one stems, 10-11 mm. in length used for each sample. R = irr2L/t.
|
Oil Injected |
Injury Control |
|||
|
Distal |
Proximal |
Distal |
Proximal |
|
|
L. . . |
1200 254 35.4 67.0 |
1032 236 47.0 38.4 |
1352 252 37.3 72.0 |
612 228 50.0 20.0 |
|
r |
||||
|
/, |
||||
|
R |
||||
rate of 20.0 R. U., hut upon injection of oil increase to 38.4 R. U. The rate of the distal hydranth is reduced from 72.0 R. U. in untreated
TABLE XI
Effect of blocking circulation by means of oil in Tubularia. Stems 8-10 mm. long isolated from long thick stems. Twenty-eight stems in each sample. R = rate using t\; I -= length of primorditim in micra; r -= radius of stem in micra; /i == time in hours from cutting to formation of primordium. Rate = ^3/hrs. -10s.
|
Oil |
Control |
|||
|
Distal |
Proximal |
Distal |
Proximal |
|
|
L |
1388 286 37.8 94.0 |
884 276 46.9 45.0 |
1320 286 37.3 91.0 |
852 268 46.0 41.7 |
|
r . . . |
||||
|
t\ hours |
||||
|
R |
||||
stems to 67.0 in injected stems. This decrease is in part caused by injection of the drop into the distal region of the stem which isolates a small portion. The average length of stem from the oil to the distal end was 2.5 mm. at the termination of the experiment.
Finally, in stems where there is very little dominance there is little effect of injection of oil in the middle of the stem. This result was obtained from some 8-10 mm. pieces cut from long, thick stems. Table XI shows that neither the proximal nor distal hydranths are affected to any extent by the injection of oil. This result is to be expected from the section dealing with the use of a middle ligature to isolate the two ends. In stems where the distal hydranth exerts little dominance it was found that little effect was produced on the rate of regeneration of the proximal hydranth by a ligature.
The experiments on injection of oil into the gastrovasctilar cavity
RATE OF REGENERATION IN TUBULARIA 173
of Tubularia shows that the oil isolates the proximal end of the stem from the distal end, producing about the same effect as a ligature. In the former case the circulation is blocked while in (he latter both circu- lation and cellular transmission are blocked. It becomes important then to see just what the oil drop does in the gastrovascular cavity. It has already been pointed out that when the drop is small it has little effect in blocking dominance. Also, when there is little dominance, little or no effect of injection of oil is found. Therefore it is safe to say that there are no toxic chemical or physical effects of the oil on the cells which come in contact with the oil.
Examination of the region into which the oil is injected shows it to be firmly held in place by the endoderm which it partially displaces. When it is not so held in place it moves during the course of hours to the distal end of the stem. Although the endoderm is displaced and perhaps the cellular connections in this layer are broken, the ectoderm remains intact and the cells can be seen to be continuous over the sur- face of the oil drop. As we were not satisfied with this observation, the conductivity over the bridge of ectoderm was tested by using an elec- trical stimulus. In some previous unpublished work on electrical stimulation in Tubularia it was found that upon applying a stimulus at the proximal end the tentacles of the distal hydranth would respond. Three stems in which dominance was blocked by injection of oil were treated in this manner and in each case the tentacles of the distal hydranth responded to an electrical stimulus applied at the proximal end. From these observations there can be little doubt that the ecto- dermal connections over the oil drop are morphologically and physio- logically intact and that dominance is not transmitted over this layer.
Discussion
Child (1907) pointed out that in Tubularia mesembryanthemum both the length of the cut stem and the level at which the stem was cut were factors determining the time for regeneration and size of the primordium. Driesch (1899) before this had measured the primordium of halves of stems and found that the oral half (distal half) formed longer primordia than the aboral (proximal half). Driesch also showed that the hydranths emerged faster in the oral (distal) half as compared with the aboral (proximal) half. Thus the regional differences in re- generative capacity in the stem of Tubularia are by no means new. The new treatment of the facts by combining two variables, size and time, into a rate has not been suggested hitherto. By utilizing both variables it is possible to express the rate of change within the stem at any level and so compare rates under various conditions. It is hoped that the
174 L. G. BARTH
rate as defined in this paper is a measure of the chemical changes in- volved in the differentiation of the hydranth from the stem after cut- ting. It is not sufficient to express these changes in terms of time only since two hydranths of different sizes may regenerate in the same time and certainly the larger hydranth must have utilized more material than the smaller. Therefore the rate of chemical change must have been higher in the region which formed the larger hydranth. Simi- larly two hydranths of the same size may require different times for regeneration.
A second difference between these experiments and those of early in vestigators is the use of a ligature to isolate regions in order to test their rate of regeneration. Driesch (18W) and Child (1^07) cut stems into halves and thirds and allowed both ends to regenerate. The size and time for regeneration of distal (oral) hydranths is modified by the presence of a second region of regeneration. Both investigators showed that the size and time factors varied with the length of stem cut. In my experiments by the use of the ligature the second regenerating region is eliminated and the size kept constant. Thus the "inherent" rate is measured. It is proposed that the "inherent " rate of regenera- tion be used as a base so that the effect of variables such as length of cut stem and the rate of regeneration of a second hydranth can be studied by means of appropriate ligatures.
Since the stem of Tubularia shows a gradient of "inherent" rates there must be graded differences in the concentration of some substance or substances, which account for these different rates. Further, be- cause the rates are higher in the younger (distal) regions of the stem it is reasonable to assume that the substances are of the nature of a synthetic factor which is able to convert available materials into a hydranth.
We will let this inherent synthetic factor be represented by E and assume that the concentration of R is proportional to the inherent rate of regeneration as measured by isolation of parts of the stem. E is then present in highest concentration in the young cells at the distal end and in lowest concentration at the proximal end. Then Fig. 2, giving rates of regeneration, may be taken to indicate the relative con- centration of E at various regions of the stem since only internal factors are responsible for these differing rates at various levels <>t the stem.
But E is not the only factor affecting rate. In Fig. 3 it was shown that increase in length of stem adjacent to the regenerating region will also increase rate of regeneration. It is evident that something from the middle of the stem travels to the ends and causes an increase in rate. Let us call this factor or substance S. S is transported through
RATE OF REGENERATION IN TUBULARIA 175
the gastrovascular cavity in the circulation which is easily observed in Tubular ia. A cross-section of the stem shows four channels in the endoderm and in the intact stem particles can be seen travelling up one side of the stem and back down the other so that a fairly rapid circulation exists. Timing the flowing particles gave a velocity of 6 mm. /minute. This means that in a short stem 6 mm. long a com- plete circulation of the contents of the coelenteron should take place in 2 minutes. If one end was using up materials in rapid regeneration it is conceivable that the concentration of substance, S, in the circula- tion would be lowered considerably so that at the opposite end sub- stances might pass into the gastrovascular cavity from the cells and so inhibit regeneration by removal of available materials.
The effect of S on rate of regeneration is difficult to measure but we may take Fig. 3, which shows increase of rate with increased volume of stem as a provisional measure of S. Obviously, however, this does not give us the effect of 5 in very low concentrations. That 5 is a very important factor is seen by comparing the inherent rate of the proximal end with the relative rate (Table IV, 14-mm. and 5-mm. stems). In these cases something is actually removed from the proximal region so that although the stem is much larger the rate of regeneration is lower when the distal hydranth is regenerating.
The situation may be summarized as follows. Regeneration is essentially the transformation of stem into hydranth and this requires
E
I chemical changes : S +±H. Let us assume E to be a catalyst in the cell
which transforms 5 into H, H being the substances necessary for hydranth differentiation. The reaction is reversible, as Child (1923) has shown that hydranths may dedifferentiate into stem. I have also observed a hydranth partially differentiate from ccenosarc and then return to coenosarc. E, as we have pointed out, is present in the cells and is in highest concentration at the distal end. 5 is in the cells and also the gastrovascular cavity and is present in greatest amount in the longest stems.
In long stems, where 5 is high (Table IV, 25-mm. and 20-mm. stems) it appears that S is partitioned to the distal and proximal hydranths according to their inherent rates of regeneration, as repre- sented by E. One way of expressing it is that, with increase in 5 while Ep (concentration of E at the proximal end) and Ed (concentra- tion of E at the distal end) remain constant, II v (concentration of //at proximal end) and Hd (concentration of H at distal end) increase pro- portionally. This result is the expected one.
1/6 L. G. EARTH
The difficulty comes when we consider short stems, where Ep (concentration of E at the proximal end as measured by the proximal inherent rate) is close to Ed (concentration of E at the distal end as measured by the distal inherent rate) yet apparently 5 is used by the distal hydranth and not by the proximal hydranth. All that can be assumed is that the factor Ed converts 5 into II as fast as it appears in the coelenteron, so that the concentration of 5" is always below the minimal value necessary for regeneration at the opposite end. Since Ed > E,, it can synthesize II at a lower concentration of S. By writing the reaction as a reversible one, we may even suggest that // — » S at the proximal end in the case of short stems. This whole thought de- pends on the assumption that the difference between Ed and Ep in short stems, although small, is great enough to lower the concentration of 5 below a minimal value, Sm for the proximal hydranth. As S in- creases in amount (in longer stems) the concentration in the coelenteron rises above the minimal value Sm for the proximal hydranth and re- generation starts at the proximal end though at a low rate at first. S thus becomes a limiting factor for regeneration in low concentrations.
It will be seen that the formal explanation given above is readily applied to other results, such as those from middle ligatures and the results showing the difference between relative and absolute rates of regeneration.
From the comparison of the absolute rate of the distal hydranth Da and the relative rate of the same hydranths Dr in Table III, it is ap- parent that there is a maximal rate of regeneration for the distal hydranth, or a maximal concentration of 5 above which synthesis of // is not increased perceptibly. This is not unlikely.
The above discussion throws the entire effect of dominance on the transport of available substances (S) for regeneration. In Tnbularia the transport system is extremely simple, and it seems an ideal system for experimentation. If in short stems the circulation of 5 is blocked (as by an oil drop) both the distal and proximal ends should regenerate independently of each other and no dominance will be exerted by the distal end. The experiments in this paper indicate that such is tin- case and thus show that S is a factor which circulates in the fluid of the gastrovascular cavity.
Summary
\ . The rate of regeneration of Tubularia has been measured by the formula R • • Trr-L/t where r • - radius of cross-section, 1. -- length ot regenerate, / = = time in hours for formation of primordium or the time in hours for emergence of the fully formed hydranth.
2. The rate of regeneration of isolated parts of the stem decreases
RATE OF REGENERATION IN TUBULARIA 177
rapidly from the distal to more proximal regions. An increase in the length of stem adjacent to a regenerating end increases the rate of its regeneration.
3. When two regenerating regions are competing they partition the effect of adding the middle region of the stem if the stem is long. If the Stem is short the distal end becomes dominant and inhibits the prox- imal end.
4. A method has been devised for measuring dominance by expres- sing it as the percentage reduction in rate of regeneration due to the presence of the distal hydranth.
5. The circulation within the stem of Tubularia has been blocked by means of injection of a drop of oil with the result that the dominance (inhibition) exercised by the distal regenerating end over the proximal end is blocked.
6. A generalized mechanism based on a synthetic factor E in the cells and a circulatory factor S is suggested as a formal explanation of the phenomenon of dominance.
LITERATURE CITED
CHILD, C. M., 1907. An analysis of form regulation in Tubularia. IV. Regional and polar differences in the time of hydranth formation as a special case of regulation in a complex system. Arch. Entw.-mech., 24: 1.
CHILD, C. M., 1923. The axial gradients in Hydrozoa. Biol. Bull., 45: 181.
CHILD, C. M., 1929. Physiological dominance and physiological isolation in de- velopment and reconstitution. Arch. Entw.-mech., 117: 21.
DRIESCH, HANS, 1899. Studien iiber das Regulations-vermogen der Organismen. Arch. Entw.-mech., 9: 103.
MORGAN, T. H., 1902. Further experiments on the regeneration of Tubularia. Arch. Entw.-mech., 13: 528.
PEEBLES, F., 1931. Some growth-regulating factors in Tubularia. Physiol. Zool., 4: 1.
CYTOLOGICAL INVESTIGATIONS OF COLPODA CUCULLUS
GEORGK \V. KIDDKR AND C. LLOYD CLAFF
(From the Arnold Biological Laboratory of Brown University and the Marine Biological Laboratory, Woods Hole, Massachusetts)
Inasmuch as the various species of Colpoda have interested many previous investigators, especially with regard to factors of encystment and excystment, it was thought worth while to make a careful study of the best known species, Colpoda cucullus Muller. So many interest- ing and hitherto unreported phenomena have been observed that it \\as thought advisable to present only the cytological results in this first report. Subsequent reports will deal with the many and interest- ing observations on other phases of the problem which are at hand and these will be supplemented by further and more complete data.
The life histories of various species of Colpoda have been investi- gated from time to time starting with the work of Stein in 1854. He reported the encystment of Colpoda cucullus, the subsequent division into two, four, eight and even sixteen smaller individuals and their ultimate escape from the ruptured cyst. Rhumbler (1888) made an extended study of the process of encystment and division in Colpoda cucullus and C. steini and distinguished between the division cysts, 'Theilungscyste," and those cysts within which division does not occur, the permanent cysts, " Dauercyste." He describes in some detail the appearance and activity of freshly encysted and dividing forms, noting that the cilia are retained throughout the division process but are lost during their stay in the permanent cysts. He observed quite accurately the gradual loss of the food inclusions, during perma- nent cyst formation, although his interpretation of the method of this loss is open to serious question.
\\Vnyon (1926) has given a rather diagrammatic account of the division of Colpoda steini, agreeing with the accounts of Stein and Rhumbler but adding some details of the nuclei. Wenyon's observa- tions were made on fixed and stained material while those of Stein and Rhumbler were mostly obtained from the living material.
Very recently Penn (1U37) has reported the occurrence ot binary and quadruple division without encystment in a strain of Colpoda cucullus. Kncystment before division occurred in his race only when the cultures were old or when the cultures became crowded. He was
178
CYTOLOGICAL INVESTIGATIONS OF COI.I'ODA CUCULLUS 1/9
able to induce the formation of cyst walls by placing "healthy in- dividuals" in infusions containing gelatinous masses of bacteria. In this report are given descriptions of the nuclear phenomena which take place during the divisional process, but details are lacking.
There has been no cytological report of the nuclei of the "per- manent" or resistant cysts published to date, as far as we are aware, and the reports of the nuclei during division have all been fragmentary. It is not difficult to see why resistant cysts have eluded observation in the past when one considers the fact that most nuclear investigation has been done after the employment of the various haematoxylin stains. As pointed out by Goodey (1913) the cyst walls ("ectocyst and endo- cyst") stain intensely with both Heidenhain's and Delafield's haema- toxylins. We have found it impossible to study the contents of the resistant cysts after treatment with these standard stains and no doubt previous workers have experienced like difficulty. It is with the aid of the nucleal reaction developed by Feulgen and Rossenbeck (1924) that the details of the nuclear complex of resistant cysts have been rendered observable.
Haematoxylin stains do not react on the division cyst wall as they do on the resistant cyst walls, mainly because of the comparative delicacy of the former. It is possible to obtain a rough idea of what happens to the nuclei during division by employing these stains. But because of the densely packed food inclusions in the cytoplasm during this period fine details are obscured. After the Feulgen reaction, however, the history of the nuclear components may be readily fol- lowed. It has been found that this history is a rather surprising one and one that may shed considerable light on the role of the macro- nuclear chromatin in ciliate metabolism. Therefore, because we have observed with considerable exactitude the nuclear phenomena both during division and during the stay within the resistant cyst, and because we feel that these observations will contribute to our under- standing of related phenomena among ciliates in general, we offer the following cytological report as the first one of a comprehensive nature on this common organism, Colpoda cucullus.
MATERIAL AND METHODS
Colpoda cucullus is a very common form and may be collected in a great variety of places. Our original material wras obtained from dry hay taken from the banks of a brackish stream near Stuart, Florida. The hay was placed in spring water from which enormous numbers of the ciliates were later collected and transferred to small dishes. From these dishes a number of motile specimens were selected with a micro- pipette and placed in individual isolation culture dishes in a drop of
180 GEORGE \V. KIDDER AXI) C. I.LOYD CLAFF
twenty-four-hour-old diluted hay tea. After a number of encystments and segregations had taken place one motile ciliate \vas selected and all of the others discarded. All glassware was then scrupulously cleaned and boiled for a long period of time to remove any danger of contamination. This obvious precaution was to insure the presence of only one species with which to work. All of our material, therefore, has descended from a single organism.
The standard culture medium used in this work, while not the only one successfully used, has given consistent results and the isolation lines grown in it have exhibited a surprising vitality. It consists of nothing more than ordinary hay tea, used after twenty-four hours at room temperatures. Into this medium single motile ciliates were placed and invariably at the end of twenty-four hours three to four divisions had occurred. All that was necessary to obtain resistant cysts was to allow multiplication to proceed for forty-eight hours or longer without adding fresh medium. Under those conditions the division rate decreased until finally all of the ciliates secreted the heavy wall characteristic of the resistance phase, accompanied by the other phenomena described below. Because of the high fission rate, the hardiness of the species and the predictable response to certain environ- mental conditions there was always an abundance of material in every phase of the life history under investigation.
For the preparation of permanent slides for cytological study we employed special "recovery dishes"1 for the collection of material designed by one of us (C. L. C.). Small circular flat-bottomed Pyrex dishes with straight sides, measuring 23 mm. X 12 mm. (inside measurements), were prepared. These dishes will accommodate Y% inch circular cover glasses, allowing them to rest on the bottom with very little free space about the edges. Into one of the dishes was placed a cover glass, in some cases thinly coated with fresh egg albumin. The dish was then filled with culture medium and inoculated with a single motile Colpoda. By frequent microscopic examinations the stages desired could be accurately noted, the cover glass taken out and placed in the fixing fluid. In this way we were able to recover hundreds of the different stages on each cover glass. Another ad- vantage this method has is that the organisms do not tend to pile up but each adheres to the glass more or less separately. There is a minimum of debris which makes for clarity of the final preparations. The success of these "recovery" dishes caused us to give up entirely our earlier methods of concentrating by centrifuging and of selecting individuals under the dissecting microscope.
1 These dishes, known as the ClalT Recovery Dishes, are being put on the mar- k' t by Clay-Adams Company, Inc., New York City.
CYTOLOGICAL INVESTIGATIONS OF COLPODA CUCULLUS 181
As mentioned in the beginning, our clear preparations resulted from the use of the Feulgen nucleal reaction. We made many prepara- tions with the various haematoxylins and carmines, and while these preparations were entirely satisfactory for the motile forms and proved very useful, they were mediocre for divisional phases and entirely use- less for resistant cysts. We found that certain modifications of tin- standard Feulgen procedure were advantageous. After much experi- mentation it was found that the following times gave excellent results and we highly recommend them for future work on Colpoda cucullus: acid hydrolysis — 15 minutes at 60° C.; fuchsin sulphurous solution— 4 to 5 hours; sodium bisulphite-hydrochloric acid wash — 15 minutes (at least three changes). It was found that the increased time of washing had a decided tendency to clear the cytoplasm of any trace of free fuchsin and rendered the preparations beautifully clear.
Many types of fixing fluids were used but it was found that wher- ever cyst walls were present the more penetrating varieties gave, as would be expected, the best results. Therefore most of our material was fixed in Schaudinn's with 5 per cent acetic, corrosive sublimate in 95 per cent alcohol with 5 per cent acetic acid added, or the Gilson- Carnoy mixture. The latter fixing fluid gave the best results on the resistant cysts.
Our Feulgen prepared material was very often counterstained with either fast green in 95 per cent alcohol, methylene blue in 70 per cent alcohol, or the acid component of the Borrel mixture (Calkins, 1930). The last-named stain was modified as to balance from the original, containing 1/3 indigo carmine to % picric acid. Counter stains reacted well in the cytoplasm of trophic and divisional forms, penetrating the thin cyst membrane of the latter with ease. The resistant cyst mem- branes appear to be entirely impermeable to fast green, methylene blue and the indigo carmine component of the Borrel mixture. The picric acid of the Borrel stain penetrates very readily, however, staining the cytoplasm an intense yellow. After this counterstain striking preparations are obtained if the material has been taken from a culture in which trophic forms, divisional cysts and newly formed resistant cysts are present. The cytoplasm of the trophic forms and the divi- sional cysts is a brilliant green while the cytoplasm of the resistant cysts is yellow.
It should be mentioned that numerous preparations employing the silver nitrate method of Klein were made on the trophic forms as an aid to our positive determination of the species at the beginning of the investigation. The extreme variability of the size and form of these ciliates is so great that specific identification becomes difficult without
182 GEORGE \Y. KIDDKR AND C. LLOYD CLAFF
careful study of the ciliary pattern. The question of variation of the ciliary pattern \ve hope to present in another report.
It has been found possible to check very accurately the times when the various nuclear changes occur during each phase by starting with resistant cysts, inducing them to excyst and fixing material at frequent, timed intervals. In carrying out these timed observations a large number of " recovery dishes" were prepared at one time and their con- tained organisms fixed in order, first noting the condition of each in the living state. Due to this procedure we have a complete series of slides demonstrating the nuclear activity from excystment, through several reproductive divisions and through a second resistant encystment. As a result of these observations we are certain that the sequence of events to be described is the normal one and occurs in a regular fashion in this ciliate. All the cytological details observed from the timed material have been repeatedly checked from mass cultures.
THE LIFE CYCLE — GENERAL
Conjugation has never been observed by us in our strain of Colpoda cucullns and as far as we have been able to determine it has never been reported in the literature. Enriques (1908) mentions conjugation as having been noted by him in Colpoda steini but not in C. cucullns. Therefore we will use the term "life cycle" to denote the sequence of events taking place from one resistant cyst stage back to another.
Colpoda cucullns appears to possess the ability to accommodate itself to a wide variety of environmental conditions. The great factor in this extreme accommodation seems to be its ability to secrete, in a minimum of time, protective or semiprotective cyst membranes. Under normal conditions all reorganizational phases accompanying reproduction and resistant encystment are carried out when the organ- ism is enclosed in some type of membrane or membranes. Feeding, only occurring in the trophic stage, is accomplished in a very efficient manner, for within a few minutes after resistant excystment the cyto- plasm will be found to be literally packed with spherical food vacuoles.
\\ lien placed ill fresh culture medium the resistant cyst undergoes certain changes very rapidly. These changes lead to a rupturing of the heavy outer cyst wall (ectocyst of Goodey, 1913) and finally to liberating the swimming ciliate from the thin inner wall (endocyst of Goodey). One of us (C.L.C ) has been investigating the mechanism of excystment under normal and experimental conditions and will report these findings at an early date.
The newly excysted ciliate is devoid of food vacuoles but very rapidly engulfs great quantities of bacteria and bacterial debris. This
CYTOLOGICAL INVESTIGATIONS OF COLPODA CUCULLUS 183
food is invariably formed into spherical compact masses and distri- buted at random throughout the cytoplasm. One average sized ciliate (70 11) may have as many as two hundred such food vacuoles (Plate I, Figs. 1 and 2). Growth ensues, the ultimate size appearing to be dependent on the excellence of the cultural conditions but independent of the ability to reproduce. Within a few hours after excystment the trophic form begins to round up, without, however, losing the motility of the cilia. The cytostome becomes indistinct and a thin membrane is secreted outside the still moving cilia. This is the "Theilungscyste" of Rhumbler (1888) and undoubtedly corresponds to the "pseudocyst" of Tillina canalifera recently described by Turner (1937). In the vast majority of cases quadruple division occurs within this cyst resulting in the liberation of four small daughter ciliates. The cilia may be continually observed throughout the process. Occasionally binary division occurs resulting in the liberation of but two daughter ciliates in which case they are proportionately larger in size. We have never observed divisions resulting in more than four daughters although Stein (1854) has reported as many as sixteen daughter ciliates escaping from a division cyst. The report of Penn (1937), in which he maintains the normal condition in Colpoda cucullus to be quadruple division without encystment, was not confirmed in our study. Binary and quadruple division without encystments were encountered only occasionally among the great number of encysted forms irrespective of the bacterial condition of the medium. It is possible that lack of encystment is a peculiarity of the strain studied by Penn.
If no additions are made to the culture medium for a period of forty-eight hours the rate of division cyst formation decreases and stops and the ciliates become very small and less active. They round up very quickly and secrete a heavy, wrinkled cyst wall about them- selves. This results in the resistant cyst which is able to withstand desiccation. The formation of resistant cysts may be induced by simply drawing off a part of the medium and replacing it with old medium from a culture in which resistant cyst formation has taken place. In the literature are numerous accounts of the necessity for drying before resistant cyst formation occurs (Barker and Taylor, 1931 ; Bodine, 1923; etc.). We have found that resistant cysts form readily even in an abundance of fluid, and may be kept for long periods without recourse to evaporation. As pointed out by Penn (1937), if conditions inducing resistant cyst formation occur rapidly enough (evaporation, according to Penn), some of the division cysts become resistant cysts by the simple expedient of forming a heavy wall about the outside of the division cyst wall. Thus resistant cysts may contain one, two
184
GEORGE \V. K I ODER AND C. LLOYD CLAFF
or occasionally four cells, all motor organelles dedifferentiating until nothing can be seen but the nuclear apparatus and the granular cytoplasm.
A
B
K
TEXT FIG. 1. Diagrammatic representation of the "life cycle" of Colpoda cuculhis illustrating the sequence of events during reproduction and resistant cyst formation. A-J, normal reproductive activity repeated (/ to B) under favorable cultural conditions. K-O, resistant cyst with space between N and () representing the lapse of an indefinite amount of time while the arrow from A' to N represents a short space of time during which macronuclear reorganization and chromatin elimi- nation takes place. Arrow from O to A represents the return of favorable conditions for excystment.
A young resistant cyst possesses many food inclusions but these are rapidly absorbed, leaving characteristic refringenl inclusions in their -if, id. U'ilhin a few hours even the refringent bodies have disappeared
CYTOLOGICAL INVESTIGATIONS OF COLPODA CUCULLUS 185
and the cytoplasm becomes compact and evenly granular, containing the nuclear apparatus. The compactness of the cytoplasm is brought about by the actual shrinkage in the size of the organism even after the cyst wall has been laid down. Resistant cysts are always very much smaller than all but the smallest of the trophic forms or division cysts.
The above account agrees essentially with the previously published descriptions of the "cycle" of Colpoda cucullus and will serve as a basis for the details of the nuclear activity described below.
Text figure 1 illustrates the normal course of events during the "life cycle" of Colpoda cucullus which, with the accompanying legend, will serve to clarify the above description.
THE NUCLEI OF COLPODA CUCULLUS
Enriques (1908) gives as a specific characteristic of Colpoda cucullus the possession of a macronucleus with a lobed karyosome "cariosoma lobato." He was able to observe this after staining with carmine. From his figures it may be assumed that, except for the karyosome, the macronucleus is devoid of stainable material (chromatin). Wen- yon (1926) figures the macronucleus of Colpoda cucullus with irregular karyosome-like bodies toward the center. He does not state what stain was employed but it would be safe to assume the use of a hxma- toxylin. That both of these observations were due to the type of stain used is demonstrated by a comparative study of organisms stained in borax carmine, Heidenhain's haematoxylin and after the Feulgen reaction. In the first two cases the stainable material seems to be concentrated toward the center of the macronucleus surrounded by a faintly stained, irregular periphery. But after the Feulgen re- action the picture is reversed. The chromatin is seen as irregular plaques surrounding and extending into the colorless nucleoplasm (Plate I, Fig. 1). After any of the counter stains used with the Feulgen reaction the nucleoplasm is sharply contrasted to the chromatin and is seen to have the form described as the karyosome by previous workers. Penn (1937) illustrates this general condition in his figures from Feulgen prepared material. We wish to emphasize the chromatin configuration of the macronucleus of the trophic form, therefore, as being in the form of irregular plaques around the periphery and sur- rounding the non-chromatin nucleoplasm of the center, because of the prevalent use of this character in classification (see the description of Kahl, 1931).
A single micronucleus is always present in the trophic stage. It lies near the macronucleus but may be on any side of it. The chro- matin is quite compact and in fixed preparations is usually seen to have
186 GEORGE W. KIDDER AND C. LLOYD CLAFF
shrunken away from the nuclear membrane. The micronucleus is quite commonly elongated, flattened on one surface, and slightly pointed at the ends. \Ye have never found more than the single micronucleus in the trophic forms. Penn (1937) mentions the oc- casional occurrence of two or four micronuclei in his strain. \Ye believe that it is possible that he interpreted the balls of extrusion chromatin from the macronucleus as micronuclei inasmuch as he completely over- looked these interesting bodies (see our Plate I, Fig. 23).
In our material the micronucleus behaves in an orthodox fashion during the divisional activities. After the division cyst membrane is laid down it enlarges and assumes a spherical shape. The chromatin becomes slightly less compact and is seen to be finely granular (Plate I, Fig. 3). It then becomes distinctly striated and the whole nucleus elongates and enlarges. The striations within the chromatin become more marked and finally irregular threads may be observed all oriented with the long axis of the nucleus (Plate I, Fig. 4). Contraction of the chromatin threads results in the formation of the metaphase plate (Plate I, Figs. 5, 6, 9). There appear to be a great many chromosomes and we were never able to make even an approximate count, as a glance at the figures will reveal. The anaphase is formed by a separa- tion of the chromosomes of the metaphase plate into two groups; the nature of this separation has not been determined. The two daughter chromosome groups move to opposite ends of the fully formed spindle leaving between them definite clear fibers which remain into tin- elongated telophase (Plate I, Figs. 7 and 8). The chromatin becomes compact again in the daughter telophase groups and these move farther apart (Plate I, Figs. 10 and 19), retaining for some time the connecting strand formed from the nuclear membrane. Finally this strand breaks and the two micronuclei round up and take their positions at opposite ends of the now elongated macronucleus (Plate I, Figs. 11 and 12). This sequence is repeated prior to every cell division without any appreciable variation.
It is within the chromatin of the macronucleus that events occur that offer an interesting problem in ciliate cytology. Before the division cyst membrane is laid down the chromatin loses its plaque-like configuration and becomes flocculent, being roughly dispersed through- out the nucleoplasm. It is still slightly granular but the granules are exceedingly minute (Plate I, Fig. 2). By the time the division cyst membrane is formed the chromatin of the macronucleus has begun to take on a definite configuration (Plate I, Fig. 3), that of granular aggregates suspended in the clear nucleoplasm (Plate I, Figs. 9, 10 and 11). This configuration is retained until the daughter organisms are
CYTOLOGICAL INVESTIGATIONS OF COLPODA CUCULLUS 187
ready to emerge from the cyst, in most cases through the two divisions. The chromatin aggregates stain uniformly and rather intensely but because of their scattered condition the macronucleus as a whole appears as a loosely knit body, lying among the numerous food inclu- sions. By the time the micronucleus has reached its telophase stage the macronucleus has begun to elongate. This elongation marks the future division plane of the cell, being always at right angles to it. Further elongation stretches the macronuclear membrane until a typical constriction appears separating the chromatin into two daugh- ter halves (Plate I, Fig. 12). The two daughter halves of the macro- nucleus quickly separate and round up and the fission plane forms, dividing the cell into two equal daughter cells, each containing a single micronucleus and macronucleus and numerous food inclusions (Plate I, Fig. 13).
Shortly after cell division there becomes differentiated simultane- ously within each daughter macronucleus an irregular, granular mass of chromatin. This mass stains much more intensely than the general macronuclear aggregates and is first seen lying close to the nuclear membrane. It is rapidly separated from the chromatin aggregates and is pushed out from the surface of the macronucleus being sur- rounded by a pocket formed from the macronuclear membrane. This activity is nearly or exactly synchronous in each daughter macro- nucleus (Plate I, Fig. 14). This deeply staining chromatin mass is the "extrusion chromatin" or "residual chromatin" and is ultimately broken away from the macronucleus and cast into the cytoplasm (Plate I, Figs. 15 and 17). Within the cytoplasm it becomes com- pacted into an intensely staining ball which rapidly diminishes in size until it disappears from view. It is usual that complete absorption of the extrusion chromatin is accomplished before the start of the second fission.
The second division is initiated, as in the first, by the activity of the micronucleus. All stages appear to be the same as in the preceding division (Plate I, Figs. 18 and 19) with the result that four diuighter ciliates are formed within the original division cyst membrane. Oc- casionally the cyst wall becomes soft and irregular and liberates the two daughters before the second division. Figure 17 represents a case of binary fission just before the liberation of the daughter cells. The extrusion chromatin has not been absorbed yet, and may not be until after the daughters become free-swimming organisms.
Following the second cell division each macronucleus again under- goes a reorganizational process whereby more extrusion chromatin is formed and cast out into the cytoplasm (Plate I, Figs. 20 and 21).
KS8 GEORGE W. KIDDER AX I) C. I.I.OVI) CLAFF
After this process has taken place the four daughter ciliates become more active, the cyst wall becomes softer and more irregular and finally ruptures. In the majority of cases enough time elapses during the freeing process for the extrusion chromatin to become absorbed but sometimes the cyst membrane ruptures early and the young daughter organisms each carry with them the remains of the residual ball (Plate I, Fig. 23). This residual ball might easily be mistaken for a supernumerary micronucleus, as we feel sure has been the case in the report of Penn ( 1937) on the occurrence of more than one micronucleus.
When, as occasionally happens, as mentioned above, binary or quadruple division takes place without the formation of a definite cyst membrane there occurs exactly the same nuclear activity as found in the normal division cyst. Chromatin extrusion follows each division in as regular and predictable a fashion as that just described. Figure 22 represents the end result of a quadruple division without cyst formation. These daughter ciliates are completely reorganized and all trace of the residual chromatin has disappeared. We have been able to find all stages representing the above process of chromatin elimina- tion in these divisions but because of the duplication of the conditions found during the normal process and because division without cyst formation was the exception in our material we felt that illustrations would be redundant.
The above process of the alternation of a free-swimming, feeding organism with reproduction within the division cyst is repeated every eight hours, on the average, either in isolation or mass cultures so long as fresh medium is provided. Whether or not there will be found a waning vitality over extended periods of time we cannot tell at present. Experiments testing this point are being carried out and will be re- ported at a later date.
When conditions within the culture change due to the accumulation of waste products resistant cysts are formed. This process is accom- plished in a very short period of time and involves the laying down of a thick, relatively impermeable wall, the absorption of the food bodies and the concentration of the cytoplasm. No activity on the part of the micronucleus is observed but the macronucleus proceeds to the most profound reorganization as yet reported for any holotrichous ciliate.
Resistant cyst formation in a trophic organism proceeds with a diminution in size and a rounding up of the ciliate. As the heavy cyst wall is secreted the cilia disappear and the cyst becomes firmly attached to the substrate. The nuclear apparatus is much the same in appearance as in the trophic stage except for the tact l licit both
CYTOLOGICAL INVESTIGATIONS OF COLPODA CUCULLUS 189
nuclei become smaller and slightly more dense (Plate II, Fig. 24). The food inclusions are rapidly absorbed and the cytoplasm then con- tains numerous minute refringent bodies (Plate II, Figs. 25 and 26). After the disappearance of the food balls but before the disappearance of the refringent bodies the macronucleus begins to elongate and to differentiate into two distinct regions. One end becomes more in- tensely staining than the other as if the chromatin was becoming com-
EXPLANATION OF PLATES
All figures are of Colpoda cucullus, and with the exception of Figs. 4-8 the magnification is X 1,000. Figs. 4-8 represent a magnification of X 2,000. All figures are from preparations treated with the Feulgen reagent. Fixatives used were Schaudinn's fluid with 5 per cent acetic acid for the preparations illustrated on Plate I and Gilson-Carnoy fluid for the illustrations on Plate II. The cilia, which are present in all stages represented on Plate I, have been omitted from the illustra- tions. Asterisks (*) represent residual chromatin.
PLATE I Explanation of Figures
FIG. 1. Trophic ciliate showing typical nuclear apparatus, food inclusions, contractile vacuole and cytoplasmic vacuoles. Note the plaque-like arrangement of the macronuclear chromatin.
FIG. 2. Ciliate about to undergo encystment prior to reproduction. The macronuclear chromatin has become dispersed and flocculent.
FIG. 3. Early division cyst. Micronucleus enlarged and the macronuclear chromatin beginning to collect in aggregates.
FIGS. 4-8. Representative micronuclei during mitosis.
FIG. 9. Micronucleus in metaphase and macronuclear chromatin in the form of irregular aggregates.
FIG. 10. Later stage. Macronucleus elongating.
FIG. 11. Daughter micronuclei at the poles of the elongated macronucleus.
FIG. 12. Constriction of the macronucleus.
FIG. 13. Plasmotomy completed and the two daughter macronuclei rounded up. No indication of chromatin differentiation for elimination as yet.
FIG. 14. Budding off of extrusion chromatin (*).
FIG. 15. Slightly later stage.
FIG. 16. About the same condition as the previous stage and a timed prepara- tion. This cyst represents the first one formed after emerging from the resistant cyst. Note small size.
FIG. 17. Probably a case of binary fission with the two daughter ciliates about to leave the wrinkled cyst membrane. Note extrusion chromatin (*).
FIG. 18. Prophase of the second division. The extrusion chromatin has been absorbed in the cytoplasm.
FIG. 19. Constriction of the macronuclei for the second division.
FIG. 20. Second cell division completed. Extrusion chromatin being given off from each daughter macronucleus.
FIG. 21. Four small ciliates about to emerge from the division cyst. Within the cytoplasm of each will be seen the residual ball of chromatin (*).
FIG. 22. A case of quadruple division without encystment. This represents the last stage with the daughter ciliates completely reorganized and about to separate.
FIG. 23. A free-swimming ciliate just after emerging from the division cyst. The residual ball of chromatin (*) has not been absorbed as yet.
190
GEORGE W. K1DDER AND C. LLOYD CLAFF
* *
> '* '*
I
:
. * .
>*H7
m?
**-
^
•.»
i '-*• "
y •
8
CYTOLOGICAL INVESTIGATIONS OF COLPODA CUCULLUS 191
. . g V^-»
•% . ife :••:-. <"
20
,
—
17
^
19
t U
16
®:.*>&:
.!&
22
192
GEORGE \V. K1DDER AND C LLOYD CLA1-T
'&
I 3
:>
24
25
26
28
29
30
31
32
35
•*
'37
»
"38
>
39
CYTOLOGICAL IN\ HSTK iATIONS OF COLPODA CUCULLUS 193
parted. This compacted region varies in size from one-third to one- half the whole nucleus (Plate II, Fig. 27). Very rapidly the compacted area buds off from the rest of the macronucleus leaving its chromatin in an irregular granular condition (Plate II, Fig. 28). The compact bud moves farther away from the nucleus until the connection between the two severs and the deeply-staining chromatin rounds up in the cytoplasm (Plate II, Figs. 29, 30 and 31). The amount of chromatin extruded varies considerably in the different cysts. In extreme cases it forms a ball as large as the remaining reorganized macronucleus but usually it is somewhat smaller (Plate II, Figs. 31 and 32). The ulti- mate resorption of the cast-out chromatin takes place by a gradual diminution in size but no apparent diminution in staining capacity (Plate II, Figs. 33 and 34). Within a few hours after encystment the extruded chromatin has disappeared and there is no further activity within the cyst until excystment. The refringent bodies have gradu- ally disappeared during this process of nuclear reorganization and the resting cyst then possesses a very clear, slightly granular cytoplasm in which the micro- and macronucleus are embedded (Plate II, Fig. 35).
PLATE II
FIG. 24. Newly-formed resistant cyst characterized by the possession of food spheres.
FIG. 25. Later stage showing the diminution of food bodies and the concomitant appearance of the small, refringent bodies.
FIG. 26. The food bodies have entirely disappeared and the refringent bodies are numerous.
FIG. 27. Early stage in the differentiation of the extrusion chromatin within the macronucleus. The extrusion chromatin stains more intensely than the chroma- tin which is destined to remain.
FIGS. 28-30. The budding off of the ball of extrusion chromatin.
FIG. 31. The connection between the extrusion bud and the macronucleus has broken. Note the globular condition of the large extrusion mass (*).
FIG. 32. A large extrusion mass (*) somewhat later than the preceding stage.
FIGS. 33-34. The extrusion chromatin mass (*) diminishes in size. Note also that the refringent bodies within the cytoplasm have disappeared.
FIG. 35. A reorganized resistant cyst with compact clear cytoplasm and a smoothly granular, compact macronucleus. In this condition the resistant cyst will remain until the cultural conditions are altered sufficiently to induce its excystment.
FIGS. 36-39. Chromatin elimination from the macronuclei of two-cell resistant cysts duplicating, in each cell, the conditions seen in the single-celled resistant cyst.
FIG. 40. Two-celled resistant cyst in which cyst formation occurred before the extrusion chromatin of the divisional reorganization was completely absorbed. The divisional extrusion chromatin is represented by the small, deeply staining balls (*d) while the resistant cyst extrusion chromatin is represented by the large masses (V).
FIG. 41. Four-cell resistant stage showing typical extrusion chromatin in each cell. This type of cyst is relatively rare.
FIG. 42. Small, clear, cyst-like structures which appeared in a few old liquid cultures and which are thought to represent degenerate resistant cysts.
194 GEORGE W. KIDDER AXI) C. LLOYD CLAFF
\\V have found a few cases where two buds of waste chromatin have been extruded from the reorganizing macronucleus but these are rare.
When, as often happens, division cysts become resistant cysts there is a complete reorganization within each of the daughter ciliates identi- cal with that of the single-celled cyst described above. Usually there appears to be enough time for the divisional reorganization of the macronucleus to proceed to completion and the subsequent re- sorption of the extrusion chromatin to take place before resistant cyst macronuclear reorganization sets in (Plate II, Figs. 36, 37, 38, 39 and 41). Rarely are there found resistant cysts where both divisional extrusion chromatin and resistant cyst extrusion chromatin are present (Plate II, Fig. 40). Also resistant cysts containing two cells are much more frequent in occurrence than those containing four cells.
DISCUSSION
In our opinion the actuality of macronuclear reorganization involv- ing the elimination of residual chromatin may possibly be demon- strated universally among the holotrichous ciliates. The definite estab- lishment of the elimination of residual chromatin as a single ball during divisional reorganization has been made in the following forms: Kid- deria (Conchophthirius} my till (Kidder, 1933a), Ancistruma isseli (Kidder, 1933/;), Conchophthirius anodontx, C. curtus, C. magna ( Kidder, 1934), Myxophyllum (Conchophthirius} steenstrupii (Rossolimo and Jakimowitsch, 1929), Allosphxrium convexa (Kidder and Summers, 1935) and Urocentrum turbo (Kidder and Diller, 1934). A number of other species undoubtedly fall into this group if we can judge by the published reports (see Loxocephalus, Behrend, 1916 and Enpoterion pernix, MacLennan and Connell, 1931). Post-divisional chromatin elimination, i.e. the casting into the cytoplasm of residual chromatin from each of the daughter macronuclei after separation, is known to occur in IclUhyophthirius multifiliis (Haas, 1933), Colpidium colpoda, C. campylum and Glaucoma scintillans (Kidder and Diller, 1934), Chilodonella labiata and C. faurei (MacDougall, 1936).
In all the cases cited above the chromatin to be eliminated is dif- ferentiated within the macronucleus prior to its division and thus ad- vertises itself. All species, therefore, in which no differentiation into regions occur prior to fission have been described as having a "clean" macronuclear division. As the vast majority of ciliates divide without encystment, and the two daughter cells separate immediately after fission, this phase of their cytology has been neglected. Ii serins entirely possible to us that if attention were paid to the young daughter
CYTOLOGICAL INVESTIGATIONS OF COLPODA CUCULLUS 195
cells after fission the occurrence of macron uclear reorganizational processes would be discovered in a great many if not all species. An- other possibility that suggests itself is that the many cases of chromatin- like fragments in the cytoplasm so often reported in ciliate studies may be explained by some process of macronuclear elimination during reorganization. \Ye wish to emphasize the necessity for more thor- ough and critical cytological work with this problem in mind to deter- mine whether or not macronuclear reorganization with chromatin elimination will be found to be a universal principle applicable to till holotrichous ciliates.
As to the exact significance of this regular though complicated process, we are still unable to say. It has been suggested (Kidder, 1933a, 19336, 1934; Kidder and Diller, 1934) that the eliminated chromatin represents worn-out material and the process might be a cleaning out of the macronucleus toward a perfect organization. It was further suggested (Kidder and Diller, 1934) that the profound reorganization which occurs at every division of Colpidium colpoda, C. campylum and Glaucoma scintillans might account for the fact that conjugation rarely occurs in these species, the reorganizational process serving to restore the cells to their fundamental condition and thereby decreasing the necessity for conjugation. The above suggestion seems to us to apply as well as any other to Colpoda cucullus.
Concerning the drastic reorganization that occurs immediately after the resistant cyst is formed, it seems logical to suppose that this represents the ridding of the macronucleus of materials no longer needed in the state of decreased or suspended metabolism. Materials accumulated in the macronucleus through the very activity of encyst- ment may be detrimental to the resting cell or to the process of excyst- ment, to come at a later date. Unfortunately we have very few records of what goes on within the resistant cysts of the various species of ciliates with which to make comparisons. We know from a few sources (see Tittler, 1935) that a process of endomyxis sometimes accompanies encystment, whereby the old macronucleus is completely discarded and a new one formed from micronuclear material. These cases would seem to represent simply a different method for accom- plishing the same general result as occurs in Colpoda cucullus, the production of a purified macronucleus.
That the reorganizational process occurring within the resistant cyst takes the place of the divisional reorganization of the macro- nucleus is denied by direct observation. In the very first division after emergence from the resistant cyst the normal reorganizational chro- matin elimination occurs. This was determined by timed preparations
196 GEORGE W. KIDDER AND C. LLOYD CLAFF
and Figure 16 illustrates a first division cyst. The small size is usual as the ciliates emerging from the resistant cysts are very small and usually reproduce before full growth is attained, a condition noted in the case of Tillina ma°na by Gregory (1909). The time factor may play an important role here, however, as there is the possibility of the aging of the macronucleus during its long period within the resistant cyst, resulting in the necessity for reorganization immediately upon again taking up an active life.
The actuality of a regular and predictable macronuclear reorganiza- tion with the elimination of quantities of chromatin during division and within the resistant cyst has been established for Colpoda cucidlus but a completely satisfactory explanation of its significance awaits further investigation. Experiments are now under way which, it is hoped, will throw some light on this question.
SUMMARY
1. A complete description of the nuclear activity of Colpoda cucullus Muller is given for the first time.
2. In our strain the normal method of reproduction takes place within a thin cyst membrane. Usually two divisions result giving rise to four daughter organisms which break out of the cyst and repeat the process. Occasionally binary fission occurs within the cyst. Rarely quadruple division occurs without encystment, as described by Penn (1937).
3. Following each cell division there occurs a reorganizational pro- cess within the daughter macronuclei resulting in the elimination of a quantity of residual chromatin. The residual chromatin is cast into the cytoplasm where it is absorbed. Elimination of residual chromatin is regular and synchronous in each cell whether the division has occurred within a cyst or not.
4. When cultural conditions are poor resistant cysts are formed.
5. The resistant cysts are formed by the secretion ot a heavy cyst membrane, the absorption of the food inclusions and the concentration of the whole protoplasmic mass.
6. Immediately following the formation of the resistant cyst membrane the macronucleus undergoes a profound reorganization during which a variable, but always a considerable amount of chromatin is budded off and cast into the cytoplasm where it is absorbed. \<> micronuclear activity occurs at this time.
7. The question of chromatin elimination from the macronuclei of holoirichous ciliates is discussed and the opinion expressed that this phenomenon may represent a universal principle.
CYTOLOGICAL INVESTIGATIONS OF COLPODA CUCULLUS
LITERATURE CITED
BARKER, H. A., AND C. V. TAYLOR, 1931. A study of the conditions of encj imrni
of Colpoda cucullus. Physiol. Zool., 4: 620.
BEHREND, K., 1916. Zur Conjugation von Loxoct-phalus. Arch. f. Protist., 37: 1. BODINE, J. H., 1923. Excystation of Colpoda cucullus. Jour. Expcr. Zool., 37: 115. CALKINS, G. N., 1930. Uroleptus halseyi Calkins. II. The origin and fate of the
macronudear chromatin. Arch.f. Protist., 69: 151. ENRIQUES, P., 1908. Sulla morfologia e sistematica del genere Colpoda. Ar<li.
Zool. Ex per. ft Gen., 8: (N. & R.) 1. FEULGEN, R., AND H. ROSSENBECK, 1924. Mikroscopisch-chemischer Nacluvcis etc.
Zcitschr. Physiol. Chem., 135: 203. GOODEY, T., 1913. The excystation of Colpoda cucullus from its resting cysts, and
the nature and properties of the cyst membranes. Proc. Roy. Soc. London,
(B) 86: 427. GREGORY, L. H., 1909. Observations on the life history of Tillina magna. Jour.
Exper. Zool., 6: 383. HAAS, G., 1933. Beitrage zur Kenntnis der Cytologie von Ichthyophthirius multi-
filiis Fouq. Arch.f. Protist., 81: 88.
KAHL, A., 1931. Die Tierwelt Deutschlands. 21 Teil: Protozoa. Fischer, Jena. KIDDER, G. W., 1 933<;. Studies on Conchophthirius mytili de Morgan. I. Mor- phology and division. Arch.f. Protist., 79: 1. KIDDER, G. W., 19336. On the genus Ancistruma Strand (Ancistrum Maupas).
I. The structure and division of A. mytili Quenn. and A. isseli Kahl.
Biol. Bull., 64: 1. KIUDER, G. W., 1934. Studies on the ciliates from fresh water mussels. II. The
nuclei of Conchophthirius anodontae Stein, C. curtus Engl., and C. magna
Kidder, during binary fission. Biol. Bull., 66: 286. KIDDER, G. W., AND W. F. DILLER, 1934. Observations on the binary fission of
four species of common free-living ciliates, with special reference to the
macronuclear chromatin. Biol. Bull., 67: 201. KIDDER, G. W., AND F. M. SUMMERS, 1935. Taxonomic and cytological studies on
the ciliates associated with the amphipod family Orchestiida? from the
Woods Hole district. Biol. Bull., 68: 51. MACDOUGALL, M. S., 1936. Etude cytologique de trois especes du genre Chilo-
donella Strand. Morphologic, Conjugaison, Reorganisation. Bull. Biol.
France et Belg., 70: 308. MACL.ENNAN, R. F., AND F. H. CONNELL, 1931. The morphology of Eupoterion
pernix, gen. nov., sp. nov. : a holotrichous ciliate from the intestine of
Acmaea persona Eschscholtz. Univ. Calif. Publ. Zool., 36: 141. PENN, A. B. K., 1937. Reproduction in Colpoda cucullus. Arch.f. Protist., 88: 366. RHUMBLER, L., 1888. Die verschiedenen Cystenbildung und die Entwicklungsge-
schichte der holotrichen Infusoriengattung Colpoda. Zeitschr.f. wissensch.
Zool., 46: 549.
ROSSOLIMO, L. L., AND FRAU K. JAKIMOWITSCH, 1929. Die Kernteilung bei Con- chophthirius steenstrupii St. Zool. An?., 84: 323.
STEIN, F., 1854. Die Infusionsthiere auf ihre Entwicklung untersucht. Leipzig. TITTLER, I. A., 1935. Division, encystment and endomyxis in Urostyla grandis
with an account of an amicronucleate race. La Cellule, 44: 189. TURNER, J. P., 1937. Studies on the ciliate Tillina canalifera n. sp. Trans. Am.
Microsc. Soc., 56: 447. WENYON, C. M., 1926. Protozoology. New York.
EMBRYONIC DETERMINATION IN THE ANNELID, SABELLARIA VULGARIS
I. Tin-: DIFFERENTIATION OF ECTODERM AND KNDODEKM \\'HEN SEPARATED THROUGH INDUCED EXOGASTRULATION
ALEX B. NOVIKOFF
(From the Department of Zoology, Columbia University, and the Marine Biological Laboratory, Woods Hole, Mass.)
INTRODUCTION
The method of transplantation has brought to light the abilities of certain embryonic parts to induce, or organize, the development of other embryonic structures in adjacent tissues. The use of this tech- nique has, however, only recently been extended to the so-called "mosaic' eggs. Tung (1934) reported experiments involving the fusion of eggs at the two-cell stage, fusion of micromeres and macro- meres, and rotation of micromeres in the eggs of Ascidiella. He found no deviation from the normal in the development of the cells giving rise to notochord and muscle. However, there was evidence that the sense organ and the ectodermal adhesive organ were induced. Two years later, there appeared abstracts by Horstadius (1936) and by the writer (1936) describing the development of transplanted blastomeres of Cere- bratulus and- Sabellaria respectively. Horstadius (full report, 1937) found no evidence of induction in fusions of various quartets of the 16- cell stage. The essentially similar results of the writer will be described in greater detail in the second paper of the present series.
It is the purpose of this first paper to describe the results of an experiment which, though similar to isolation experiments, possesses certain of the added advantages of the transplantation method. The experiment consists of the production of larva? in which the endoderm cells grow out, during gastrulation, from their normal position under- lying the ectoderm. This affords a way of testing the independent differentiation of the two layers and, by inference, the effect of one layer upon the other during normal development.
MATERIAL AND METHODS
The egg used is that of the marine annelid, Sabellaria vulgaris. That this is an egg of the classical mosaic type is shown by the com- pletely partial development of isolated blastomeres (Halt, 1932). The
198
DETERMINATION IN SABELLARIA 199
procedure for obtaining fertilized eggs of this species has been described elsewhere (Novikoff, 1937). The larva? of this_ species are closely similar to those described by Wilson (1929) for the two European species, alveolata and spinulosa.
Exogastrulae can be produced in considerable numbers by a treatment which removes the vitelline membranes of fertilized ova. A variation of the method employed by Hatt (1931) has proved to be quite effective. At any time before they have begun to change their form in preparation for cleavage, the eggs are washed once with an isotonic NaCl solution brought to pH 9.6 by the addition of Na2CO3 (.98 gram in 1 liter of solution) and are then placed into 5 cc. of the same solution. The eggs adhere to each other in this medium, forming large aggregates, but as the egg membranes disappear, the aggregates break up. When this has occurred, 5 cc. of acidified NaCl solution (0.45 cc. 1.0 N HC1 in 100 cc. 0.53 N NaCl) are added, bringing the medium to pH 8.2. When the eggs have settled to the bottom of the dish, they are washed once in sea water and transferred to a Syracuse dish containing fresh sea water. Among the swimming larvae which develop, the exogastrulse can readily be distinguished from the normal larvae in the dish.
Representative stages in the development of exogastrulae and control larvae are described in the text. All text figures are camera lucida drawings of living larvae; the description accompanying each figure refers not to that specific larva but is a general description of larvae at that stage in development. The time intervals given are approximate, for temperatures ranging from 19° to 24° C.
DESCRIPTION OF RESULTS
It is possible to follow the details of ectodermal and endodermal differentiation in living larvae. In the normal course of development the ectoderm gives rise to the following structures: a prototroch of long active cilia, at first encircling the animal but later incomplete on the dorsal surface; chromatophores, which are yellow when they form but later change to green; an apical tuft which disappears when re- placed by short non-motile cilia; paired chaetae-sacs from which extend long serrated bristles; a single, short, non-motile posterior cilium; several long, non-motile dorsal cilia in the region of the prototrochal gap; a single orange-red eye spot on the left side, towards the dorsal surface; and rapidly moving cilia ("neurotroch") in a ventral depres- sion leading to the mouth cavity of the fully formed trochophore. The endoderm differentiates into a tripartite gut consisting of oeso- phagus, stomach, and intestine. The ccsophageal cells are lined in-
200 ALEX B. NOVIKOFF
ternally with many long actively-moving cilia. The stomach and in- testine are also ciliated but the motion of their cilia differs from that of the oesophagus; they move more slowly and remain more rigid during the effective stroke.
In order to compare the development of the eggs from which the membranes have been removed with that of the normal eggs, a series of stages in the development of both are described. Those stages have been chosen which show definite advances in structural differentiation.
The early cleavages of the membraneless eggs are similar in all respects to those of normal eggs. Although the details of the later cleavages have not been described for this species, any marked differ- ence in this process between the two types of eggs would have been noted. But no such variations were apparent. The first noticeable departure from normal development appears at the time when the larva? show the first signs of movement. This is usually between six and seven hours after fertilization and probably corresponds to the time when gastrulation is normally occurring.
Comparison of Complete Exogastrulx with Normal Larvx
Six to Seven-hour Larva. — Normal (Fig. 1.4). There is no regularity in the shape of the larvae at this time; it may vary from a spherical to a roughly conical form. The vitelline membrane, retaining the many wrinkles characteristic of the earlier stages in development, is still far removed from the body of the larva. Through the membrane, there project a group of long cilia at the future anterior end of the larva, and many short, rapidly moving cilia which form a narrow girdle completely encircling the larva.1 These cilia constitute the apical tuft and the prototroch respectively. The larva is rather opaque and it is with difficulty that the large internal cells can be distinguished. Directed towards the inner cells and situated slightly posterior to the prototroch is the blastoporal indentation.
Exogastrula (Fig. IB). These larvae are more uniformly regular than the normal larvae of the same age. Not obscured by the vitelline membrane as they are in the normal larvae, the cilia of the apical tuft and prototroch are more easily distinguished. The distance between the apical end and the prototroch (i.e., the pretrochal region) is the same as it is in the normal larva, but the post-trochal region is con- siderably lengthened by a posterior outgrowth of cells. No blastopore is present.
Thirteen-hoiir Larva. — Normal (Fig. 1C). The larva has now rounded out and there are only slight irregularities in the surface.
1 The cilia in the early larvie may be bright out more clearly with darkfield illumination.
DKTKKMINAT10N IN S.\ UKI.I .A I' 1 \
201
It has elongated somewhat in the antero-posterior direction and is just beginning to assume the typical trochophore appearance. The apical tuft is directed forward as the larva swims, rotating on its longitudinal axis. The prototroch, which formed a complete ring about the animal at the preceding stage, is interrupted by a short region with no cilia; this gap marks the dorsal surface of the trocho- phore. At the posterior, broader end there projects a fine non-motile cilium. The surface of the larva now has a general gold-yellow colora- tion when viewed by transmitted light. Through the outer ectoderm cells the "anlage" of the gut is discernible. From a crescent-shaped cavity at the apical end, a narrow canal leads through the gut cells to open on the surface at the posterior end. The blastopore, which has become progressively smaller up to the eleventh hour of develop- ment, has completely disappeared.
FIG. 1. A. Normal larva. Seven hours after fertilization. Shows apical tuft and prototroch. Posterior to the prototroch is the blastoporal indentation. B. Exogastrula. Seven hours. Viewed from posterior end. Shows apical tuft, prototroch, and elongate post-trochal region. C. Normal larva. Thirteen hours. Shows apical tuft, prototroch, posterior cilium, and differentiating gut. D. Exo- gastrula. Thirteen hours. Shows apical tuft, prototroch, posterior cilium, and elongate post-trochal region.
Exogastrula (Fig. \D}. This larva is considerably more elongate, antero-posteriorly, than the normal larva of the same age. The structure of the apical tuft, prototroch, and posterior cilium is normal, and the apical tuft is in the usual location. On the other hand, the relative positions of both the prototroch and the posterior cilium are strikingly different. The former is situated more anteriorly in the larva; the latter is to one side of the larva and projects laterally instead of posteriorly. The general yellow coloration of the embryo does not extend very far beyond the equator; the cells at the posterior end are colorless. The crescent-shaped cavity and the longitudinal canal present in the endoderm cells of the normal larva at this stage are not present.
202
ALEX 15. XOYIKOFF
Eighteen-hour Larva. — Normal (Figs. 2 A and B}. Since the pre- ceding stage, the larva has become more spherical and more regular in outline. The membrane, almost free of irregularities, is more closely applied to the larval surface. The cilia of the apical tuft are now considerably longer and they move slowly in two distinct groups. The prototrochal gap has grown in size; the posterior cilium is si ill present, and the yellow pigment has become localized into a number of larger areas on the surface of the body. The endoderm cells within are now assuming the form of the gut; one can recognize the three main divisions — the ventral oesophagus runs anteriorly into the wider stomach, which in turn is continuous with the intestine. The only lumen visible in the gut is the narrow canal described in the earlier larva; it runs through the stomach and intestine and opens as the anus. Slightly anterior to the prototroch and on either side of the
FIG. 2. Larvae, eighteen hours after fertilization. A. Normal larva. Side view. Shows apical tuft, prototroch, gut, and pigmented areas. 11. Normal larva. Dorsal view. Shows prototrochal gap, stomach and intestine portion of gut, and posterior cilium. (.'. Exogastrula. Post-trochal region is elongated and hollow.
stomach there has developed a spherical structure, the chueta-sac, from which the long bristles later grow.
Exogastrula (Fig. 1C}. Changes similar to those described for the normal larva have occurred in the apical tuft, prototroch, and pigmen- tation of the exogastrula. The apical tuft is composed of two groups of Ion- cilia. The prototrochal gap extends over the dorsal surface. The yellow pigment has become more localized, but since the pigment was originally restricted to the more anterior part of the larva, the posterior portion is completely colorless. The difference between this larva and the normal one of the same age is, however, much more pronounced than at the preceding stage. In contrast to the fairly spherical form of the normal, it is narrow and long. It possesses no internal gut and in the posterior half there appears a large cavity. The posterior cilium is again laterally displaced.
DKTKKMIX. \TION IX S M'.KI I \i:l \
203
Twenty-four-hour Larva. — Normal (Figs. 3A and B). Important changes have occurred between this and the preceding stage of develop- ment. In the apical region there have appeared, alongside the apical tuft, a number of short, non-motile cilia. On the dorsal surface of the larva, in the region of the enlarged prototrochal gap, there are two or three long non-motile cilia. The posterior cilium is still short and non- motile. The large pigmented areas have been broken up into well- defined yellow chromatophores. The gut has acquired a continuous lumen, lined with rapidly moving cilia. The cilia of the oesophagus differ from those of both the stomach and intestine in that they are slightly longer, are more numerous, per unit surface area, and move more rapidly. In their movement, they are bent more than are the stomach-intestinal cilia. The oesophagus opens to the exterior through
B
FIG. 3. Larvae, twenty-four hours after fertilization. Abbreviations: at, apical tuft; ac, apical cilia; pc, posterior cilium; dc, dorsal cilia. A. Normal larva. Ventral view. Mouth, oesophagus, and intestine visible. The pigmented areas are broken up into well-defined chromatophores. Two pairs of bristles are present. B. Same larva as A. Side view. C. Exogastrula. Shows elongate, hollow, post-trochal region. On the surface, toward the posterior end, short cilia are visible.
the mouth, the intestine through the anus. From each of the chaetae- sacs there now extend two bristles. The position and structure of these bristles are very characteristic. Arranged at intervals along their length are serrated "collars." As the larva swims about, the bristles, projecting posteriorly, are seen being rotated about the long axis of the animal. That muscles have developed can be deduced from the observed contraction of the larva? and the movement of the chaetae-sacs and bristles.
Exogastrula (Fig. 3C). At this stage, the exogastrulae can easily be distinguished, even with low magnification, from the normal animals. The most striking feature is the behaviour of the bristles when the animal is in motion. The larva rotates as usual as it moves in a forward direction. But the bristles project laterally instead of
204 ALEX B. NOVIKOFF
posteriorly, and they are accordingly turned over and over in a wide arc, rather than rotated on the longitudinal axis of the larva. Mewing the animal from above, the bristles therefore appear and disappear from sight. No gut is present, and the cavity of the preceding stage has increased noticeably in size. On the surface of the posterior portion of the larva there are now rapidly beating cilia, of the same size as the cilia in the gut of the normal larva. They are usually found in two distinct extensive areas, with the larger of the two areas covered with cilia characteristic of the normal oesophagus. It is possible to distinguish the single asymmetrically-placed posterior cilium. The larva* show the following changes over the preceding stage, which are similar to the changes in normal development: (1) the appearance of short, non-motile cilia at the apical end; (2) the formation of -•mailer, yellow chromatophores, although in the exogastrula- they are
FIG. 4. Larv.p, I UPIII y-nine hours afier I'crl ili/at ion. Abbreviations as in Fig. 3. A. Normal larva. Side view. Mouth, stomach, and intestine are well- formed. The bristles on the left side only have birn drawn. B. Exogastrula (slightly flattened). Dorsal view. Only four of the eight bristles have been drawn.
limited to little more than the anterior half of the larva; (3} the >!>••. el< .| .HP-HI oi i]i ii al i ili.i near i In- c.'.ip in i lie pn »totroch; ; I i In- appearance of contractions of the larva, albeit these contractions are considerably weaker than those of the normal larva; (5) the formation of two pairs of bristles of the same notched structure- as described lor the normal.
Twenty-nine-hour Larva. — Normal (Fig. 4^4). The larva is begin- ning to assume the appearance of the fully-developed trochophore. The membrane is closely applied to the surface of tin- larva. The apical tuft has disappeared, and only the non-motile cilia remain at the anterior end. The prototroch is raised slightly, forming the so- called "hood fold" i Wilson, 1(J29). The single posterior cilium per- sists. The small yellow chromatophores are taking on a green tinge. The stiff dorsal cilia are longer and more numerous than in 1 he previous
DETERMINATION IN SABELLARIA
205
stage. The oesophagus, lined with many active cilia, leads into the stomach. The stomach is partially separated from the intestine by an incomplete shelf; the intestine is lined with cilia of the type found in the stomach. The number of bristles varies from three to four pairs. On stimulation (i.e., when the slide or the dish is tapped gently), the larva rapidly contracts and the bristles are spread into a fan-like arrangement. After several seconds in this position, the animal returns to its usual form.
Exogastrula (Fig. 4.B). The general appearance of the larva is similar to that of the preceding stage. The three or four pairs of bristles project laterally and, as the animal moves, are rotated in the manner described above. On stimulation, they are spread out as in
FIG. 5. Larvae, forty-eight hours after fertilization. Abbreviations as in Fig. 3.
A. Normal larva. Dorsal view. To the left of the stomach is the eye spot (e).
B. Exogastrula. Dorsal view. A large part of the post-trochal region is covered with short cilia.
the normal larva, although to a lesser extent. There is no internal gut, and a large cavity is present inside the larva. The chsetae-sacs project into the cavity. As in the normal larva of this age, only the apical cilia persist at the anterior end. The dorsal cilia have elongated and increased in number. The chromatophores, limited to about the anterior half of the larva, are becoming greener in color. The cilia on the exterior of the post-trochal region are found in two regions. Of these only the ventral region has the cilia of the cesophageal type.
Forty-eight-hour Larva. — Normal (Fig. 5A). This is a well-devel- oped trochophore. The post-trochal region has lengthened somewhat. The apical cilia are numerous and the number of dorsal cilia has also increased. There is a single posterior cilium. A second, posterior row
206 ALEX B. NOVIKOFF
of short cilia has appeared on the hood fold, just posterior to the long prototrochal cilia. On either side of the mouth there has developed a fold covered externally with cilia, the lip fold. The chromatophores have a more pronounced green coloration. On the left side of the larva, near the stomach, there has appeared a single eye spot, com- posed of closely packed orange-red granules. The gut is clearly differentiated into oesophagus, stomach, and intestine. The stomach has increased in size and has become more spherical. Running posteriorly from the mouth, in a groove along the ventral surface, are the long cilia which constitute the neurotroch. A larva at this time ordinarily has four pairs of bristles, which project posteriorly.
Exogastrula (Fig. SB). Changes similar to those occurring in the normal larva? have taken place between this and the preceding stage. The apical cilia have become more numerous. The prototroch is better developed and a row of short cilia has been added posterior to it. The dorsal cilia have increased in number and si/e, and the posterior cilium is single and dorsal. On the dorsal surface, anterior to the prototroch, may be seen the eye spot. Four pair of bristles project laterally as do the bristles of the earlier larva1. There is no internal gut. The cilia characteristic of the oesophagus and the intestine, together with the cilia of the neurotroch and the lip fold, cover the post-trochal region almost completely. Special mention must be made of the eye spot, since it was not seen in all the larvae examined. It could not be found in thirteen of the ninety-three larva? studied.
Partial Exogastrulx
There are always found in the culture dishes containing the mem- braneless larva1 a few larva- which are intermediate in structure between those described as complete exogastrulu* and the normal. These larvae are illuminating for an understanding of the structure of the more extreme type. Figure 6 shows two views of a well-developed trochophore of this type. Not all the structures are visible in the drawings, but the larva possesses every structure found in the normal animals, ah hough these structures have in some cases differentiated in abnormal positions. The most outstanding feature of these larva? is the gut structure. The stomach and intestine are clearly normal in form, position, ,md type of ciliation. The intestine opens posteriorly, as usual, through the anus. At the anterior end of the stomach there is a circular opening lined with cilia which move in a manner characteristic of the cilia of the o^sophageal-stomach opening in the normal larva. But this opening leads here not to the oesophagus, but to the outer surface of the larva. In fact, there is no internal
DETERMINATION IN SABELLARIA 207
oesophagus at all. On the ventral surface of the larva is a long, conical outgrowth, completely covered with long actively-moving cilia which are characteristic of the oesophagus.
DISCUSSION OF RESULTS
There are two important points of difference between larvae which develop from eggs without membranes and normal larvae. The first is the complete absence of an internal gut and the second the striking deviation from the normal form of the larva. Instead of possessing the fairly spherical shape, they have elongate bodies in which the posterior structures, the posterior cilium and the paired bristles, are dorsally displaced. The distortion of the posterior part of the larva may be accounted for if we assume that the endoderm cells, which differentiate normally inside the ectoderm to form the gut, grow out so that they come to lie on the surface of the larva.
B
FIG. 6. Partial exogastrulae. Seventy-two hours after fertilization. A. Dorsal view. The opening to the stomach, the stomach, and the intestine are visible. B. Ventral view. The bristles have been omitted in the drawing; only their points of origin in the chsetae-sacs are indicated. The opening to the stomach may be seen at the base of the long conical projection covered with short cilia.
That this posterior outgrowth of the endoderm cells has indeed occurred is shown by the following considerations: 1. The gut is com- pletely lacking in these larvae, and within the post-trochal region of the larva there is a large cavity. 2. The cells situated at the posterior end of these larvae are not pigmented and never develop chromato- phores. In the normal larva the chromatophores are found in the ectoderm; only the endoderm is colorless. 3. In the normal larva, motile cilia do not appear on the external surface of the post-trochal region until approximately thirty-five hours after fertilization. \\ hen they do develop, they are restricted to a narrow region, the neurotroch. But in the membraneless larva cilia begin to appear on the surface of the post-trochal region before the end of the first day of development. By twenty-four hours after fertilization they have covered two wide areas. That these cilia are not those of the neurotroch is clear from
208 ALEX B. NOYIKOFF
the fact that they cover two extensive regions and that they develop twelve hours earlier than do the neurotrochal cilia. Their appearance coincides precisely with the development, normally, of cilia on the inner surface of the endoderm cells. It is also possible to detect differences in the cilia which correspond to the differences between the resophageal cilia and the cilia of the stomach and intestine.
\Ye may therefore conclude that in these eggs the endoderm has been turned inside out. Whatever the mechanism involved, the end result of this process is comparable to the exogastrula1 that have been described in the sea-urchin (Herbst, 1893; Driesch, 1893) and in the amphibia (Holtfreter, 1933a, b).
In the partial exogastrulae, the stomach and intestine develop normally, and only the o-sophagus is exogastrulated. And, as we would expect on the basis of our assumption, the general form of the larva is more like the normal than is that of the complete exogastrulu.
The most significant feature of exogastrula development in Sabel- laria is the complete self-differentiation of both endoderm and ecto- derm. In the amphibian egg, Holtfreter found that normal differen- tiation of the ectoderm is dependent upon contact with the mesendo- derm. When the latter, instead of coming to lie beneath the ectoderm as it does normally, evaginates, it leaves the ectoderm a wrinkled, hollow sac in which no trace of differentiation into nervous tissue appears. On the other hand, the mesendoderm, although turned inside-out, undergoes self-differentiation; it produces gut, thyroid, pancreas, liver, notochord, musculature, kidney, and gonad. Thus, in the normal course of development, the endoderm differentiates independently of the ectoderm, but the ectoderm must be in contact with the endoderm, or gut roof, to differentiate normally. It is this gut roof which acts in the capacity of an organizer — which induces the formation of a nervous system in the ectoderm.
In Sabellaria, on the other hand, there is no deviation from normal differentiation of either the ectoderm or endoderm in the exogastrula?. The endoderm cells retain their morphological polarity, so that they develop cilia on what is in these larvae their external surface. And although the tripartite nature of the gut is lost, the type of cilia devel- oped by the exogastrulated <rsophagus cells differs from that of the stomach and intesiine in the same way as do the normal cesophageal cilia from the normal stomach-intestinal cilia. The ectoderm, too, differentiates normally. The pigment develops and concentrates into yellow areas, which Liter break up into smaller chromatophores and become distinctly green in color. The apical cilia form at the anterior end ;uid the apical tuft disappears. The serrated bristles develop
DETERMINATION IN SABELLARIA 209
normally, although the cells from which they arise have been displaced to one side (dorsally). On the dorsal side there develops the orange- red eye sppt. Muscles, as evidenced by contractions of the larva, are also differentiated.
Thus, we find no evidence of any inducing effect of the endoderm upon the ectoderm, or vice versa, in the development of Sabellaria. There may possibly be some question concerning the self-differentiation of the eye spot in the exogastrula since in 13 out of 93 cases no eye spot was observed. In more than one hundred normal control larvae studied, none was found without the eye spot. However, when individual untreated eggs are isolated in small drops of sea water beneath cover-slips it is found that they develop into larvae in which only the eye spot fails to develop. In these isolates, it is clear that the absence of the eye spot is not due to the morphological absence of an inducing region below the ectoderm. The failure of the ectoderm to form the red pigment may perhaps be due to some general factor such as increased pressure or lower oxygen tension. The same is probably true of those exogastrulae in which no eye spot is seen.
Thus, the absence of the eye spot in some of the exogastrulae does not constitute a serious objection to the conclusion that there is no inducing ability of either the ectoderm or endoderm, as tested by their effects upon one another during development. It is possible that in the exogastrulae the ectoderm has not been completely isolated from the endoderm. The cells which will give rise to the two layers are in contact with each other at one point. Through the study of exo- gastrulae alone one can not exclude the possibility of diffusion of materials from one part of one layer across cell boundaries to any part of the other layer; nor can it be denied that the cells may exercise in- ducing effects upon each other during cleavage. These criticisms merely point to the limitations of this mode of approach. To overcome them, another method may be used — that of transplantation of blasto- meres. The results of such experiments in the egg of Sabellaria vulgaris form the subject-matter of the second paper in this series.
SUMMARY
1. A simple method for the production of complete and partial exogastrulae in Sabellaria vulgaris is described.
2. In the complete exogastrulae, no internal gut is formed. In the partial exogastrulae there is no internal oesophagus, but the stomach and intestine develop normally.
3. Both the ectoderm and endoderm show complete self-differentia- tion in the exogastrulae, indicating that one layer exercises no inducing
210 ALEX B. NOVIKOFI-
effect upon the other between six and forty-eight hours of development , from gastrulation to the well-developed trochophore.
It is with pleasure that the writer expresses his gratitude to Professor L. ( i. Barth for untiring assistance and constant encourage- ment throughout the course of this investigation.
BIBLIOGRAPHY
DRIESCH, HANS, 1893. Entwicklungsmechanische Stuclien. VII, Exogastrula und
Anenu-ria. Mittheil. Zool. Station zu Neapel, 11: 221. H.VTT, Pn KKI , 1931. La fusion expcrimentale d'oeufs de "Sabellaria alveolata L."
et leur developpement. Arch, de Biol., 42: 303. HATT, PIKK.RE, 1932. Essais experimentaux sur les localizations germinales dans
1'oeuf d'un Annelide (Sabellaria alveolata L.). Arch. d'Anat. Micros., 28:
81. HEKUST, CURT, 1893. Experimentelle I'ntersuchungen iiber der Flinfluss der
veranderten chemischen Zusammensetzung des umgebenden Mediums auf
die Entwicklung der Thiere. 2 Theil. Mittheil. Zool. Station zu Neapel,
11: 136. HOLTFRETER, JOHANNES, 1933</. Die totale Exogastrulation, eine Selbstablosung
des Ektoderms voni Entomesoderm. Arch.}. Entii'., 129: 669. HOLTFRETER, JOHANNES, 1933/>. Organ isierungsstuf en nach regionaler Kombination
von Entomesoderm mil Ektoderm. Biol. Zentralbl., 53: 404. HORSTADIUS, SVEN, 1936. In\x'>-t igation> on determination in the early development
of Cerebratulus (Abstracl I. Ki»l. Hull., 71: 406. HORSTADIUS, SVEN, 1937. Experiments on determination in the early development
of Cerebratulus lacteus. Biol. Bull., 73: 317. NOVIKOFF, ALEX B., 1936. Transplantation of the polar lobe in Sabellaria vulgaris
(Abstract). Anat. Rec., 67: (Supplement 1) 57. NOVIKOFF, ALEX B., 1937. Sabellaria vulgaris. Culture Methods for Invertebrate
Animals, by (ialtsolT et al. Comstock Publishing Co. Inc., p. 187. TuNCi, Ti-Cnow, 1934. Recherches sur les potentialtes des blastomcres chez
Ascidiella scabra. Arch. d'Anat. Micros., 30: 381. WILSON, I)OU<;LAS P., 1929. The larva; of the British sabellarians. Jour. Mar.
Hiol. Ass'n., 16: 221.
EMBRYONIC DETERMINATION IN THE ANNELID, SABELLARIA VULGARIS
II. TRANSPLANTATION OF POLAR LOBES AND BI.ASK >MKKES
AS A TEST OF THEIR INDUCING CAPACITIES
ALEX B. NOVIKOFF
(From the Department of Zoology, Columbia University and the Marine Biological Laboratory, Woods Hole, Massachusetts]
INTRODUCTION
E. B. Wilson (1904a) demonstrated that the egg of the annelid, Lanice, belonged to the group of so-called mosaic eggs, since isolated halves of the two-cell stage developed into partial embryos. The anterior cell produced an embryo which lacked the post-trochal region; in the embryo formed from the posterior cell there was present a nearly typical post-trochal region. Delage (1899) had previously described a dwarf embryo from an egg-fragment of the same species. That annelid eggs, generally, are of the mosaic type is shown by the experiments of Penners (1924, 1926) with Tubifex, of Tyler (1930) with Chxtopterus, and of Hatt (1932) with Sabellaria.
Wilson (1929) summarizes the evidence which indicates that there is no fundamental distinction between the mosaic and regulative types of ova. Among regulative eggs, where correlative differentiation, or embryonic induction, is most prominent, mosaic features can be found, and in mosaic eggs, there are suggestions that embryonic induction may play a part in early development. Wilson suggests that the polar lobe of such eggs as Dentalium may function as an organizing region similar to the dorsal blastoporal lip of amphibia, since only when the lobe is present does the larva develop the apical tuft and the post- trochal region. However, in the absence of transplantation experi- ments, no final conclusions could be reached.
Schleip (1929) describes a "natural experiment" in which a second polar lobe is added to the egg of Dentalium. Among giant eggs, some are found which appear to be fusions of two ova at their vegetal hemispheres. In these eggs, a single large polar lobe may be formed, which goes in its entirety into one of the cells. This leaves one c-u with no lobe, and the other with two. However, such eggs do not develop. Schleip then tried to transplant isolated polar lobes to blastomeres, but all attempts were unsuccessful.
211
212 ALEX B. NOVIKOFF
Transplantation experiments can readily be performed in the egg of Sabellaria vulgaris. In this species, large polar lobes, similar to those found in Den tali urn, are formed in the course of the first three cleavages. Although the egg is only about sixty micra in diameter, it is not difficult to remove the polar lobe or to separate individual blastomeres. Both blastomeres and polar lobes can be fused together in desired combinations and the eggs reared through a well -developed trochophore stage. In all, 247 successful transplant operations have been studied, including 80 in which either the first or second polar lobe was transplanted.
MATERIAL AND METHOD
All experiments reported in this paper were performed at the Marine Biological Laboratory, Woods Hole, Mass , during the summers of 1935, 1936, and 1937. The animals used were dredged from Vineyard Sound and the eggs were obtained in the manner described elsewhere by the writer (1937). The egg of Sabellaria vulgaris possesses a tough, wrinkled vitelline membrane which resists cutting with glass needles. In addition, within the membrane, in the perivitelline space, there is a dense jelly which makes difficult the separation of individual blasto- meres and which completely prevents bringing blastomeres together. Thus, to perform transplantations, it is necessary to first remove both membrane and jelly from the egg.
The vitelline membrane is removed from fertilized eggs of Sabellaria by treatment with an isotonic solution of NaCl, brought to pH 9.6 by the addition of Na2CO3 (Novikoff, 1938). In most cases, the eggs are thus treated within the ten minutes that elapse between the forma- tion of the first polar body and the second. After they have been washed once in sea water, the denuded eggs are placed into a Syracuse dish of freshly-filtered sea water. The dish is allowed to remain without disturbance on the stage of a dissecting microscope. Within two or three minutes, the eggs have settled and are adhering to the glass bottom of the dish. By means of fine glass needles, each egg is then lifted from the jelly which remains adherent to the bottom. This process may have to be repeated several times in order to remove the jelly completely. The jelly being invisible, its removal can best be ascertained by bringing together the individual eggs; they come into contact with each other only when the jelly has been removed. At this time the eggs are quite sticky and if allowed to adhere too long to the dish they flatten out. Since such flattened eggs do not develop normally, it is important that the eggs be lifted, at close intervals, from the bottom of the dish.
DETERMINATION IN SABELLARIA 213
The eggs are cut, free-hand, under the dissecting microscope. In order to determine the orientations of cells when fused, small spots are marked on the eggs before they are cut. This is done by bringing into contact with the surface of the egg the open end of a fine capillary tube filled with agar, in which is dissolved a vital dye such as Nile Blue sulphate. In order to fill an exceedingly fine bore, the following pro- cedure is followed. A short piece of capillary tubing is partly filled by immersing one end in a warm solution of the agar containing the dye. When the agar has cooled, that part of the tube which has no agar is heated by a microflame and pulled out to the desired width of bore. The other end of the tube is then sealed off and the microflame applied to the part of the agar nearer the narrowed end. The agar melts and moves into the free end of the tube. On cooling, the agar, in many cases, remains at the opening of the tube. The tube is brought into contact with the egg by means of a Zeiss- Peterfi micro- manipulator. The length of time during which the agar must remain in contact with the egg varies with the concentration of dye used. In some cases, where it is possible to determine the polarity of the isolated cell without previously staining a particular region, the entire cell is stained before transplantation. This is usually done before the mem- brane is removed, and a dilute solution of Nile Blue sulphate in sea water is used.
To effect the fusion of blastomeres it is only necessary to bring
them into contact with each other, after the membrane and jelly have
' been removed. It is usually sufficient to press them together for
several seconds, although this may, in some cases, have to be repeated
several times before they finally stay together.
Following the operation, the eggs generally develop into swimming larvae within five to six hours after fertilization. They can at that time no longer be left in the open dish for they soon swim to the surface, where they are quickly torn by the surface film. The procedure which leaves the least amount of surface exposed to the air and which, at the same time, is most convenient for the detailed study of the living larvae, is to allow the embryos to develop in small drops of sea water between a glass slide and coverslip. When an embryo begins to show signs of movement, it is transferred, by means of a mouth pipette (Horstadius, 193 7a), to a small drop of sea water. Evaporation from the drop is prevented by sealing the edges of the coverslip with a thin layer of vaseline.
The results to be described are based on a total of 403 experimental larvae, 156 developing from isolated blastomeres and 247 from fusions of various blastomeres. The percentages of larvae surviving are: 73
214 ALEX B. XOVIKOFF
per cent on the first day, and 51 per cent on the second day, from isolated cells; 91 per cent on the first day, and 74 per cent on the second day, from combinations of blastomeres. As described in the first paper of the present series (Novikoff, 1938), a great number of the larvae arising from denuded eggs develop into exogastrula?, in which there is no internal gut and the endoderm cells are turned inside out. Under the coverslips, all structures differentiate in the larva?, just as they would in larger volumes of sea water, with the exception of the orange- red eye spot, and, to some extent, the yellow-green chromatophores. These are frequently absent or irregular. That the failure of the eye spot and the chromatophores to develop normally is due to some general factor such as lowered oxygen tension or increased pressure, is indicated by the development of normal eggs (i.e., eggs from which the membranes have not been removed) under similar conditions. Al- though the cell arrangement and the tissues of the larvae from such eggs are manifestly similar to those of normal larva?, the eye spot does not form and the chromatophores show the same variable character as do those of the membraneless larvae.
NORMAL DEVELOPMENT
Changes in the egg of Sabellaria vulgaris at maturation and during fertilization are described elsewhere (Novikoff, in press). Both polar bodies are formed after insemination and cleavage begins approximately twenty minutes * after the extrusion of the second polar body.
In the course of the first cleavage division, a large part of the vegetal hemisphere of the egg becomes constricted in the form of a spherical lobe. Figure 1, A is a photograph of an egg at the "trefoil" stage, when the polar lobe is at its maximum size. When viewed from the side, the lobe appears to be equal in size to either of the first two blastomeres. But when viewed from the vegetal pole, it is seen to be considerably smaller. The visible constituents of the lobe cytoplasm do not differ from that of either blastomere, except that there is no spindle area in the lobe. About fifteen minutes after it first appears, the polar lobe flows into one of the blastomeres. This blastomere is the CD cell and it is now much larger than the other, the AB blastomere (Fig. 1, B). Figure 1, C shows an egg during the second cleavage, when the second polar lobe is at its maximum size. This lobe forms in the CD blastomere only, and is smaller Mian the first lobe. When it flows back into one of the daughter cells at the completion of the division, the four quarter-blastomeres consist of two equal-sized cells, A and B (the products of the division of AB), a slightly larger cell, C,
1 All time intervals are for room temperatures, varying from 19° to 25° C.
DETERMINATION IN SAI1ELLARIA
215
and a much larger cell, D (Fig. 1 , D). It is the D blastomere which has received the contents of the second polar lobe. During the next divi- sion, when the micromeres are produced, a third polar lobe, formed
B
FIGS. 1-10. All figures, except 1 and 5, are composed of camera lucida drawings of living larvae, magnified approximately 260 times. Abbreviations used are:
p prototroch at apical tuft ac apical cilia
ec cilia of exogastrulated endoderm cells pc posterior cilium dc dorsal cilium pb post-trochal bristle
e eye m mouth
FIG. 1. Photomicrographs of living eggs of Sabellaria vulgar is. A. Trefoil stage. Side view. The polar lobe is slightly out of focus. B. Two-cell stage. Blastomere CD is to the right of A B. C. Second cleavage. The second polar lobe and two of the quarter-blastomeres are in sharp focus; the other two cells are not in focus. D. Four -cell stage. Seen from the vegetal pole. The large D cell is to the right and the A cell is uppermost. E. Early trochophore larva. Sixteen hours after fertilization. Photographed with dark- field illumination; shows apical tuft and prototrochal cilia. F. Later trochophore larva. Forty-eight hours after ferti- lization; shows stomach, intestine, dark eye spot, prototrochal cilia, and post-trochal bristles.
from the D cell, flows into the D macromere, ID. This lobe is smaller than the second lobe and is more variable than the preceding lobes;
216
ALEX B. NOVIKOFF
in many cases this lobe does not become distinctly separated from the dividing D cell. The later cleavages of the egg have not been described.
A detailed description of the development of the larva is presented in the first paper of this series (Novikoff, 1938). In the normal course of development, the ectoderm gives rise to the following structures: a prototroch of long active cilia; chromatophores; an apical tuft ; non- motile apical cilia; paired chaeta>sacs from which extend long, serrated bristles; a single posterior non-motile cilium; several long non-motile cilia on the dorsal surface; an eye spot; and a neurotroch of rapidly moving cilia. The endoderm differentiates into a tripartite gut con- sisting of oesophagus, stomach, and intestine. In many ot the mem- braneless larvie, the gut is exogastrulate and the three portions can not be distinguished. Of the ectodermal structures, those most easily observed in living embryos — and therefore the ones best suited for
FIG. 2. Larva; from isolated blastomeres. A. E-PL1 larva, seventy-one hours; with internal jjut. B. E-PLl larva, fifty hours; with exo^astrulated endoderm. C. AB larva, fifty-one hours; with internal .nut. D. AH larva, fifty-six hours; \\ith exo.ya-trulated endoderm.
the present study — are the apical tuft, the post-trochal bristles, the prototrochal cilia, and the apical cilia. The apical tuft forms at about six hours after fertilization, and persists for approximately twenty hours (Fig. 1, /'."). Before it disappears, there develop at the apical end, a number of stiff cilia; these apical cilia remain throughout larval development. The prototrorhal cilia appear at about the same time as the apical tuft and they remain throughout larval development. The post-trochal bristles make their appearance toward the end of the first day of development. They increase in length and number as development progresses (Fig. 1, F).
ISOLATION EXPERIMENTS
E-TLl . An egg from which the polar lobe is removed at the treloil stage is labelled K-l'/J. The first cleavage of E-PLl differs from i In- normal second cleavage in that no polar lobe is formed. As a result,
DETERMINATION IN SABELLARIA
217
the quarter-blastomeres are equal in size. In addition, no lobe is formed at the next division. The larva? of such eggs differ from the normal larva in that: (1) no apical tuft forms; (2) no post-trochal region (including the bristles) appears; and (3) the prototrochal cilia, although of the same size as in normal larva?, are at the posterior end of the larval ectoderm (Fig. 2A, B). Although no apical tuft is present, at ap- proximately twenty-four hours after fertilization the typical non- motile cilia appear.
AB. — The isolated AB cleaves without the formation of polar lobes and gives rise to a spherical larva similar to that of E-PL1. It lacks the apical tuft and post-trochal bristles. It develops normal proto- trochal and apical cilia, although the former are situated at the posterior end of the larval ectoderm (Fig. 2C, D).
FIG. 3. Larvae from isolated blastomeres. A. CD larva, twenty-five hours. B. CD larva, eighty-five hours. C. CD-PL2 larva, fifty-six hours. D. E-PL2 larva, forty-five hours. (On the next day, the apical tuft was no longer present.) E. E-PL2 larva, eighteen hours. It is exceptional in that it possesses two apical tufts.
CD. — During the first cleavage of CD there is formed a polar lobe of the same size as the normal second lobe. After the division, it flows into the D cell. At the second cleavage, another, smaller lobe forms from the D cell. This lobe has the variable character of the normal third lobe. The early CD larvae appear to be quite normal — the prototrochal cilia are in their usual position and the typical apical tufts are formed (Fig. 3, A}. However, since the cilia and tuft are of the normal size, they are, proportionately, too large for these reduced larva?. Later, the paired bristles are formed. However, no apical cilia develop, and in many instances (17 out of 29), the apical tuft does not disappear when it does in controls (Fig. 3, B).
CD-PL2. — -If during the course of the first cleavage of the CD blastomere, the polar lobe is removed, the next cleavage occurs without the formation of a lobe. The CD-PL2 larva possesses a typical apical
218
ALKX B. XOYIKOl-F
tuft but lacks the post-trochal structures as well as the apical cilia (Fig. 3, C).
E-PL2. — It is with difficulty that the second polar lobe is removed from a whole egg. In each of the three larva? obtained, normal apical tufts and prototrochal cilia develop. (One larva possesses two apical tufts — Fig. 3, £.) The older larva? show no post-trochal structures, but do develop apical cilia (Fig. 3, D}.
BC. — Before the cells have shifted their position after the second cleavage, it is possible to divide them so that BC is separated from AD. The BC combination forms no polar lobe and gives rise to a larva which possesses the typical apical tuft and prototrochal cilia, and a few apical cilia. It lacks completely the post-trochal region (Fig. 4, B).
FIG. 4. Larva? from isolated blastomeres. A. Al> larva, forty-nine hours. 11. B C larva, fifty hours. C. ABC larva, t unit v-six hours. D. ABD larva, sixty- six hours. E. D larva, forty-seven hours. F. C larva, thirty-three hours. G. D-PL3 larva, forty-one hours.
AD. — At the first cleavage of AD, a small lobe, which has the variable character of the normal third lobe, forms in the D cell. The larva developing from this combination of cells forms no apical tuft. It possesses normal prototrochal cilia and the typical post-trochal region from which extend the bristles. Apical cilia are also present (Fig. 4, A}.
ABC. — By destroying the D cell at the completion of the second cleavage, ABC combinations are obtained. No polar lobe is formed at the first cleavage. The larva which develops possesses the typical apical tuft and apical cilia. However, no post-trochal bristles are produced (Fig. 4, C).
ABD. — Destruction of the C cell leaves the ABD blastomeres. At the next cleavage, the D cell forms a small polar lobe. The ABD
DETERMINATION IN SABELLARIA
219
larvae develop normal prototrochal cilia, but no apical tufts. Later in development, apical cilia and post-trochal bristles appear (Fig. 4, D).
C. — No polar lobes are formed during the cleavage of the C blaslo- mere. The larva developing from the C cell develops typical proto- trochal cilia and apical tuft. But neither post-trochal structures nor apical cilia are formed (Fig. 4, F).
D. — During the first cleavage of the D cell, the small variable lobe appears and passes into the macromere, \D. The D larva has typical prototrochal cilia and post-trochal bristles. But it develops no apical tuft and no apical cilia (Fig. 4, £).
D-PL3. — In some cases, the lobe formed by the D cell constricts sufficiently so that it may be removed. The larva developing from
TABLE I
Summary of larvie obtained from isolation experiments.
|
Number Operated |
Number Surviving |
Apical Tufts + ? |
Post-trochal Bristles + |
Prototrochal Cilia + - ? |
Apical Cilia + - ? |
|
|
E-PL1 |
30 |
21 |
0 21 |
0 11 |
21 0 |
14 0 |
|
AB |
36 |
25 |
0 24 |
0 18 |
23 0 1 |
11 6 1 |
|
CD . |
56 |
47 |
37 1 9 |
32 1 |
42 0 5 |
1 26 6 |
|
CD-PL2 |
8 |
5 |
5 0 |
0 4 |
5 0 |
0 3 1 |
|
E-PL2 |
3 |
3 |
3 0 |
0 2 |
3 0 |
2 0 |
|
BC |
10 |
8 |
6 2 |
1 5 |
8 0 |
6 0 |
|
AD |
6 |
3 |
0 3 |
3 0 |
3 0 |
2 0 1 |
|
ABC |
10 |
8 |
7 0 1 |
0 5 |
8 0 |
4 1 |
|
ABD |
7 |
5 |
0 5 |
4 0 |
5 0 |
4 0 |
|
C |
17 |
11 |
7 3 1 |
0 7 |
11 0 |
0 7 |
|
D . . . |
23 |
15 |
0 12 3 |
9 4 |
15 0 |
094 |
|
D-PL3 |
7 |
5 |
0 5 |
0 3 |
5 0 |
0 3 |
|
213 |
156 |
the D-PL3 lacks the post-trochal structures as well as the apical tuft and apical cilia (Fig. 4, G).
Table I summarizes the isolation experiments. All combinations of cells develop prototrochal cilia. But only those combinations which include the substance of the C cell (earlier found in the first polar lobe) develop apical tufts; only those which have the ID cell materials (and previously found in the three polar lobes) form the post-trochal region; and the apical cilia develop only when the A or B cell is present. Three exceptions, a BC larva with bristles, a CD larva with apical cilia, and an E-PL2 larva with two apical tufts, will be discussed later.
220 ALEX B. NOYIKOFF
Behavior of Isolated Polar Lobes
The changes in form of isolated polar lobes have been described in Dentaliiim, by \Yilson (1904&) and in Ilyanassa, by Morgan (1933, 1935). In Dentalium, the isolated lobe usually constricts periodically to form lobe-like structures by a process which simulates the form.it ion of polar lobes in the- whole egg. There 'are three such constrictions, and they occur at approximately the same time as do the cleavages of the whole egg. One case is described in which the final constriction gives rise, not to a temporary lobe, but to a distinctly separated portion of cytoplasm. Wilson interprets the first of these changes as the formation of lobes within the isolated polar lobes, at the time when tin; whole egg normally forms lobes. The final stage he regards as a
PLl
PLz
V
PLl
PL 2
Fl«. 5. Form changes in isolated polar loin--. PLl, first polar lobe. PL2, second polar lobe. Numerals 1 1 indicate the time of the first four cleavages of the whole egg. Explanation in text.
permanent division of the isolated lobe into two at the time of the fourth cleavage of the whole egg, in which (he material of the polar lobe no longer forms a temporary polar lobe, but is permanently cut off by a cell division. Morgan argues against a literal interpreta- tion of the form changes in the isolated lobe as lobes. In Ilyanassa, he finds that : ( 1) The changes in the lobe are not strictly synchronous with the cleavage of the whole egg. (2) The constrictions in isolated lobes come and go at least three times, whereas in the whole egg, the lobe would appear only once more. (3) The change in shape does not, strictly speaking, give a reduced picture of the changes in the whole egg; and, (4) the form changes in the lobe resemble more the process of
DETERMINATION IN SABELLARIA 221
micromere constriction rather than lobe formation. Therefore, Morgan interprets the later changes of the isolated lobe as related to the constriction of micromeres, and not to the formation of polar lobes.
First Polar Lobe. — Isolated first polar lobes of Sabellaria were observed continuously for two to three hours. Although, as both Wilson and Morgan found, the behavior is somewhat variable, most lobes show a remarkable constancy in their changes. Of the 31 iso- lated lobes studied, 25 were of the type drawn in Fig. 5; the other six were more or less variable.
No change in the shape of the isolated lobe occurs at the time of the first cleavage of E-PL1, i.e., at the time of the second cleavage of the whole egg (Fig. 5, b). About ten to fifteen minutes after the first cleavage of E-PL1, the lobe is deeply constricted to form a de- finite lobe-like structure (Fig. 5 , c) . At the time of the second cleavage, the lobe is spherical (Fig. 5, d), but another deep constriction appears about ten to fifteen minutes later (Fig. 5, e). As this second constric- tion disappears, a slight flattening of the lobe occurs (Fig. 5,/). This lasts for a short time, and when E-PL1 cleaves for the third time, the lobe is again spherical (Fig. 5, g). A third constriction forms after the third cleavage of E-PL1 (Fig. 5, h), but before it disappears completely, the entire cell becomes irregular in outline, and long irregular pseudo- podia are formed (Fig. 5, i). These are later withdrawn (Fig. 5, j) and the cell again becomes spherical (Fig. 5, k). However, it remains in this condition for only a short time, until the irregular pseudopodia are formed again (Fig. 5, /). The extension and retraction of the pseudopodia is not synchronous with cleavage, and continues until the lobe cytolyzes. In several cases, the process was still going on twenty-eight hours after the removal of the lobe; in one case, it was observed up to forty-eight hours after the separation of the lobe.
Second Polar Lobe. — The second polar lobes were removed from five isolated CD cells, and their behavior followed continuously for two hours. Four of the five lobes produced fairly deep constrictions twice (Fig. 5). The other produced only the first of the two constrictions; at the time when the second constriction would form, the cell elongated somewhat without constricting. The behavior of the four lobes which formed the two constrictions was fairly uniform. When CD-PL^ divides for the first time, i.e., at the time of the third cleavage of the whole egg, the isolated lobe elongates slightly (Fig. 5, b'} and within a minute is rounded out (Fig. 5, c'). A similar elongation forms again, in three of the lobes, in about eight to ten minutes after the first (Fig. 5, d'}. At about seventeen to eighteen minutes after the first cleavage, the lobe develops a constriction which persists for two
ALEX B. NOVIKOFF
minutes (Fig. 5,/'). As the constriction disappears, the lobe flattens slightly (Fig. 5, g'}. \Yithin a few minutes, the second cleavage of CD-PL2 occurs. Following the second cleavage a second constriction forms (Fig. 5, i'), and as it disappears the cell becomes irregular and gives rise to pseudopodia (Fig. 5, /). The behavior of the pseudopodia is similar to those of isolated first lobes (Fig. 5, j'-ri}.
The constrictions formed by the isolated first and second polar lobes resemble in appearance the polar lobes formed by the dividing ovum. Although these constrictions do not occur at the same time as the cleavages of the ovum, they must, in some way, be related to events taking place in the egg during division. This is brought out by a comparison of the time elapsing between successive cleavages of the lobeless egg and the constrictions of the isolated lobe, in the twenty- five cases where the first polar lobe was removed and the five in which the second lobe was removed. The average time between the first and second cleavages of E-PL1 is 20.6 minutes and that between the first and second constrictions of the isolated first lobe 18.6 minutes. The time between the second and third cleavage of E-PL1 is 24.6 minutes and that between the second and third constriction is 21.2 minutes. For the isolated second lobes, the average time between the two constrictions is 21.8 minutes, while the interval between the first and second division of CD-PL2 is 21.0 minutes. The changes in the form of the isolated lobes are apparently synchronous with the cleavages of the ovum, except that all events in the isolated lobes are pushed back by a delay in the appearance of the initial constriction. This delay is approximately ten to fifteen minutes for the isolated first lobe and about seventeen to eighteen minutes for the isolated second lobe.
On the basis that the periodic constrictions formed by isolated polar lobes are correlated with cytoplasmic changes occurring in whole eggs at the time of the cleavages in which polar lobes are formed, we would expect the isolated second lobe to form one fewer constriction than the isolated first lobe. The isolated first lobe forms three con- strictions; the isolated second lobe two. And, as would be expected, the time elapsing between the first and second constrictions of the isolated second lobe (21.8 minutes) is almost identical with the time between the second and third constrictions of the isolated first lobe (21.2 minutes). However, certain differences between the behavior of the isolated lobes and that of polar lobes of the whole egg must be noted. The whole egg forms two polar lobes after (he formation ol the first lobe, and one after the second. The isolated first lobe, however, forms three constrictions and the isolated second lobe two.
DETERMINATION IN SAP.ELLARIA
223
Although there is a decrease in size of successive lobes in the whole egg, the constrictions of the isolated lobe are all of approximately the same size. But the facts to be emphasized are: (1) that there reside in the isolated polar lobes materials which take part, independently of the nucleus or the mitotic apparatus, in reactions affecting the tension at the surface of the cell, and (2) that these reactions in isolated lobes occur synchronously (if we discount the initial delay) with events in the whole egg, or in the egg from which the polar lobe has been removed.
TABLE II
Differentiation of larvae after transplantation of polar lobes and blastomeres.
|
Type of Operation |
Num- ber Oper- ated |
Num- ber Sur- vived |
Larva- with: |
Larva with: |
Larvae with: |
||||||
|
0 Api- cal Tuft |
1 Api- cal Tuft |
2 Api- cal Tufts |
OSets of Bris- tles |
ISet of Bris- tles |
2 Sets of Bris- tles |
0 Groups of Apical Cilia |
1 Group of Apical Cilia |
2 Groups of Apical Cilia |
|||
|
A. Polar lobe transplants Transplant Host PL1 Whole egg.. AB E-PL1.. PL2 Whole egg . . AB... . |
26 35 15 G 6 |
23 32 14 5 G |
7 32 14 0 6 |
10 0 0 5 0 |
0 0 0 0 0 |
8 20 10 2 3 |
14 0 0 o 0 |
0 0 0 0 0 |
7 5 3 1 0 |
15 14 7 3 3 |
0 0 0 0 0 |
|
B. Transplantation of blas- tomeres to whole egg Transplant CD |
9 10 16 11 |
9 10 14 11 |
0 0 2 2 |
o 4 10 9 |
7 6 0 0 |
0 0 0 2 |
1 8 4 8 |
G 0 10 0 |
1 2 2 1 |
5 G 12 4 |
0 0 0 5 |
|
C |
|||||||||||
|
D |
|||||||||||
|
AB |
|||||||||||
|
C. Transplantation of blas- tomeres to E-PL1 Transplant CD |
16 20 15 4 12 3 |
16 18 15 3 11 3 |
2 6 15 1 11 1 |
11 12 0 2 0 2 |
0 0 0 0 0 0 |
2 14 0 2 9 0 |
12 0 13 0 0 3 |
0 0 0 0 0 0 |
5 5 4 0 0 0 |
9 8 9 2 3 1 |
0 0 0 0 G 2 |
|
C . |
|||||||||||
|
D |
|||||||||||
|
CD-PL? |
|||||||||||
|
AB |
|||||||||||
|
Whole egg . . |
|||||||||||
|
D. Fusions of blastomeres Combination CD, CD |
20 9 6 4 12 3 1 |
19 8 4 3 10 2 1 |
0 0 1 0 11) 2 1 |
11 7 2 2 0 0 0 |
8 0 1 0 0 0 0 |
2 1 0 2 8 1 1 |
2 1 2 0 0 0 0 |
11 5 0 0 0 0 0 |
15 7 2 2 1 0 Thret |
0 0 0 0 2 0 groups o |
0 0 0 0 f) 1 "cilia |
|
CD, D |
|||||||||||
|
CD, C. . |
|||||||||||
|
C C |
|||||||||||
|
AB, AB |
|||||||||||
|
3 AB's. . |
|||||||||||
|
5 AB's |
|||||||||||
TRANSPLANTATION EXPERIMENTS Transplantation of Polar Lobes
Transplants of first and second polar lobes were made to whole eggs at the trefoil, two-cell, and four-cell stages, to AB blastomeres, and to E-PL1. The lobes were placed at the animal pole, at the vegetal pole, and at the equator of the dividing egg. Although
224
ALEX B. NOYIKOFF
fused to the blastomeres, the lobes go through essentially the same form changes as do isolated lobes. When cleavage progresses, the blastomeres may grow over and completely enclose the lobe, or the lobe may remain at one end of the larva. The lobe is still part of the larva when the cilia appear (Fig. 6A, C). Prototrochal cilia form in all larvae, and the apical tufts develop only in those which include C cells (Table II, A and Fig. 6). At about fifteen hours after fertilixa- tion, the lobe, more or less completely cytolyzed, is extruded from the larva (Fig. 6, D}. This is generally followed by a cytolysis of a portion of the embryo, especially in those cases where the lobe was deeply embedded within the embryo. Thus, only 59 out of the 80 larva?
FIG. 6. Differentiation of larvae after transplantation of polar lobes. A. AB plus PL1 larva, llj-jj hours. The lobe (stippled) is within the larva. B. AB plus PL1 larva, 24 hours. The larva has moved away from the cytolyzed spheres of the ejected lobe. C. E plus PL1 larva, 13 hours. The lobe (stippled) is within the larva. D. E plus PL1 larva, eighteen hours. Tin- lobe has been ejected. E. AB plus PL1 larva, 45 hours. The ejected lobe has broken up into small spheres. F. E plus PL1 larva, 26 hours. G. E-PL1 larva, 50 hours. The first polar lobe was in contact with the dividing egg for five hours, at the end of which time it was removed.
survived beyond twenty-four hours. Where the larva is not great ly damaged by the loss of the lobe, differentiation progresses normally. The AB and E-PLl larva> form apical cilia but no post-trochal regions (Fig. 6, B). The larva? produced by entire eggs lose their apical tufts when apical cilia appear, and they develop post-trochal regions with typical bristles (Fig. 6, F).
In many cases, the polar lobe was stained heavily with Nile Blue sulphate, before being transplanted. The dye diffused from the lobe into the adjacent cells, so that the larval tissues acquired a pronounced blue coloration.
DETERMINATION IN SABELLAKIA
225
To avoid possible effects of the lobe cytolysis on the differentiation of the larva, the lobe was, in fifteen cases, allowed to remain fused to the dividing cells only until the embryo began to show signs of move- ment, when it was removed. The contact is long enough so that the blue color of the dye diffuses from the lobe into adjacent cells. The larvae which develop from such cells are in no essential way altered by contact with the lobe (Fig. 6, G}.
The presence of the polar lobe, fused to the larva for five hours (when it is removed), or within the larva for eleven hours (at which time it is ejected), does not induce the formation of larval structures.
Transplantation of Blastomeres to Whole Eggs
Table II, B summarizes the types of transplants made to whole eggs, at the trefoil, two-cell, and four-cell stages. The transplants
FIG. 7. Differentiation of larvae after transplantation of blastomeres to whole eggs. A. E plus CD larva, 27 hours. B. E plus C larva, 24 hours. C. E plus I) larva, 79 hours. D. E plus AB larva, 40 hours.
are in some cases stained with Nile Blue sulphate before being fused to the host. It is possible to follow the stained region through the first day of development, but beyond this time the dye is not visible. The orientation is varied, but no correlation between any particular orientation and type of development is found.
The differentiation of the transplanted AB, CD, C, or D blasto- meres does not bring about the development of additional structures in the host; those structures which are duplicated in the larva arise from the self-differentiation of the transplanted cells. Seven of the nine E plus CD larvae show two apical tufts; six (out of the seven surviving) show two sets of post-trochal bristles (Fig.- 7, A). In six of the ten E plus C larvae, two apical tufts are visible (Fig. 7, B). Ten of the fourteen E plus D larvae possess two sets of bristles (Fig. 7, C). The E plus AB larvae show no duplication of either the apical tuft or
226 ALEX B. NOVIKOFF
bristles, but five of the ten possess two distinct groups of apical cilia
Transplantation of Blastomeres to E-PLI
The polar lobes are removed from eggs at the trefoil stage, and in their place are put half- or quarter-blastomeres, or whole eggs. In other experiments, transplants are placed at the animal pole, or to the side of E-PLI . As in previous operations, the transplanted cells are first stained, either in loto, or locally, to mark their polarity. The resultant larvae are summarized in Table II, C and a few of the various kinds are shown in Fig. 8. In all larva', the host cells give rise to
FIG. 8. Differentiation of larvae after transplantation of blastomeres to E-PLI. A. (E-PLJ) plus CD lafvn, 43 hours. B. (E-PLl) plus C larva, 30 hours. C. (E-PLI) plus C larva, 48 hours. D. (E-PLI) plus D larva, 17 hours. E. Same larva as shown in D, 43 hours. F. (E-PLI) plus AB larva, 54 hours. G. (E-PLI) plus E larva, 25 hours. H. Same larva as shown in G, 91 hours.
prototrochal cilia, and in at least 36 of 55, they form apical cilia. They form no apical tuft and no post-trochal bristles. The apical tufts, post-trochal bristles, or supernumerary prototrochal and apical cilia which are present in the larva* arise only through the self-differentia- tion of the transplant.
Fusions of CD, C, and D Blastomeres
Thirty-four operations involving combinations of CD cells with CD, D, and C blastomeres, and fusions of two Cecils were performed (Table II, /; and Fig. </). Among the \(> larva- with the Cl>, CD combination, 8 show two apical tufts; in 11, only one tufl is distinct Iv
DETERMINATION IN SARELLARIA
227
visible. Of the 14 larva* surviving beyond (he first day, 2 fail to develop post-trochal bristles, 2 form one set of bristles, and 11 develop two distinct sets. There are 7 larva? with the CD, D combination. Each of these larvae produces one apical tuft, and 5 of the 7 show two sets of post-trochal bristles. ( )m> larva forms one set of bristles and another forms none. Four larvae of the CD, C constitution show the following: two apical tufts, 1 ; one apical tuft, 2; no apical tuft, 1. Two survive long enough to develop bristles; in each there is one set. Only two C plus C larva? survive; each possesses only one apical tuft and none forms post-trochal bristles except those which include D cells.
FIG. 9. Differentiation of larvae arising from fusions of blastomeres. A. CD plus CD larva, 13 hours. B. Same larva as shown in A, 59 hours. C. C plus C larva, 28 hours. E. CD plus D larva, 55 hours. D. AB plus AB larva, 50 hours.
Fusions of AB Blastomeres
At the time of the first cleavage of isolated AB cells, various combinations are effected. Of the larva? surviving, ten come from fusions of two AB cells, two from three AB's, and one from five. Five of seven of the AB, AB larvae show two sets of apical cilia, and two show one set; in the single surviving larva from the combination of three AB's, two sets of apical cilia are seen ; in the larva from the fusion of the five AB's three distinct sets of apical cilia are visible. Although there is a superabundance of cellular material, none of the larvae develops an apical tuft and none develops a post-trochal bristle (Table II, D and Fig. 9, D).
228
ALEX B. NOVIKOFF
Fusions of Whole Eggs
Two eggs are fused at the trefoil stage or at the two-cell stage. The orientation of the two eggs with respect to each other is varied as follows: (1) the animal-vegetal axes of the ova remain parallel, but the eggs are rotated to different degrees; (2) one of the two eggs is inverted so that the fusion occurs at the animal poles or at the vegetal poles of the two eggs, with the rotation of the eggs varied as in (1).
j
FIG. 10. Differentiation of larvae arising from fusions of two eggs. A. Larva, 23 hours. B. Same larva as shown in A, 52 hours. C. Another larva, 27 hours. />. Another larva, 49 hours, with one internal gut. On the outer surface is an area of ciliated cells, devoid of chromatophores. E. Another larva, 55 hours, with two internal guts.
Of the twenty-live fusions effected, 23 larvae are alive at the end of twenty hours of development, and \() at the end of thirty-five hours. Of the 23, 2 larva* show no apical tuft, 7 show one tuft, and 11 sliou two tufts. Among the 19 larva?, the post-trochal bristles fail to form in one case, 4 larva; have one set of bristles, and 14 have two sets; the apical cilia are not visible in 2 larvae, 5 show one group of cilia, and 12 show two distinct groups. In none of the larva? are there more than two apical tufts, or more than (wo sets of post-trochal bristles, or more
DETERMINATION IN SABELLARIA 229
than two groups of apical cilia. Figure 10 includes three advanced larvae of different types. In one type (Fig. 10, B), the endoderm cells of both eggs have been exogastrulated. In another (Fig. 10, E), there are two distinct internal guts, of fairly normal structure. In a third type (Fig. 10, D), only one internal gut is found. This gut has the typical tripartite character and is no larger than the normal gut. On the outer surface of the larva, a wide, delimited, area is devoid of the chromatophores characteristic of the ectoderm, and the cells of this area are covered with the rapidly-moving cilia, characteristic of the gut cells. The constitution of this area is therefore interpreted as the endoderm cells of one of the eggs. Instead of giving rise to a gut, these cells have become part of the outer covering of the larva; but in spite of their new location, they continue to differentiate as they would normally.
In two instances, three eggs are fused at the two-cell stage. The resultant larvae develop three sets of post-trochal bristles. In one case, four eggs are fused at the two-cell stage, and the larva develops four sets of bristles.
DISCUSSION OF RESULTS
By means of isolation experiments, the independent developmental capacities of early blastomeres of Sabellaria vulgaris were determined. These experiments included the usual separation of half- and quarter- blastomeres, and, in addition, the separation of other combinations of cells (AD, BC, ABD, and ABC). Also, the effect of the removal of three polar lobes was studied. Table I classifies the 156 surviving larvae, with respect to the presence or absence of the prototroch, apical tuft, post-trochal bristles, and apical cilia. Prototrochal cilia were present in all larvae. The apical tuft formed only in those larvae which included the first polar lobe and the C cell. The post-trochal bristles developed only when the three polar lobes and the \D cell were present. The differentiation of apical cilia occurred only when either the A or the B cell was present. Two of the larvae were exceptions to this conclusion: one CD larva, which developed apical cilia, and one BC larva, which developed post-trochal bristles. A third larva, of the E-PL2 type, formed two apical tufts instead of one.
Having determined the fate of the blastomeres when isolated, the behavior of these same cells when placed in contact with each other atypically, or with isolated polar lobes, was investigated. The twenty- two types of transplantation experiments, including a total of 247 larvae, are summarized in Table II. Not a single case is found in which the transplant had induced the formation of any particular structure
230
ALEX H. NOYIKOFF
in (he host. In any combination of blastomeres, apical tufts develop only when Cecils are present, post-trochal bristles are dependent upon the presence of D cells, and apical cilia form only when either the A or B cell is included.
In Table III, the larva? are classified on a different basis. The number of C, D, and A or B cells included in I he makeup of tin- individuals are compared with the number of apical tufts, post-trochal structures, and apical cilia. In general, the number of apical tufts is
TABLE III
Classification of larva; developing from transplantation experiments.
|
Number of C Cells |
Total No. of Larvae |
Number of Larva with |
||
|
0 Apical Tufts |
1 Apical Tul't |
2 Apical Tutts |
||
|
0 |
88 |
88 |
0 |
0 |
|
1 |
95 |
21 |
74 |
0 |
|
2 |
64 |
3 |
28 |
33 |
|
Number of D Cells * |
Total No. of Larvae |
Number of Larva* with |
||
|
0 Sets of Bristles |
1 Set of Bristles |
2 Sets of Bristles |
||
|
0 |
69 |
69 |
0 |
0 |
|
1 |
63 |
14 |
49 |
0 |
|
2 |
62 |
4 |
12 |
46 |
|
Number of A (or /:. Cells |
Total No. of Larvae |
Number of Larva: with |
||
|
0 Apical Cilia |
1 Group of Apical Cilia |
2 Groups of Apical Cilia |
||
|
0 |
26 |
26 |
0 |
0 |
|
1 |
128 |
35 |
93 |
0 |
|
2 |
49 |
4 |
15 |
30 |
* Two larva? had 3 D cells; both had 3 sets of bristles. One larva had 4 D cells; it had 4 sets of bri-i It •-.
determined by the number of C cells present, the number of sets of post-trochal bristles by the number of D cells, and the number of groups of apical cilia by the number of A (or B} cells. The absence of a higher correlation, particularly in the cases of the apical tuft and apical cilia, may possibly be due to several reasons: (1) The larva? are actively swimming and, especially in small, healthy individuals, such minute structures may be overlooked. (The post-trochal bristles are much more readily seen. The correlation is higher with this structure than
DETERMINATION IN SAIJKLLAUI A 231
with either of the other two.) (2) Some of the operated individuals may have been injured and the failure to form a particular structure may be a manifestation of their reduced vitality. (3) Actively swim- ming larv;e may have some of their cells torn away by the surface at the edge of the drop of water. It is in many cases not possible to deter- mine whether this has occurred. (4) The mechanical effects of the neighboring tissues may prevent the cells from giving rise to the partic- ular structure. If true, this effect might possibly be conceived as a kind of regulatory process.
The most significant feature of the transplantation experiments is that in no instance is the number of apical tufts greater than the number of C cells, nor the number of post-trochal regions greater than the number of D cells, nor the number of groups of apical cilia greater than the number of A (or B] cells.
Of special interest are several larvae which developed from a fusion of two eggs at the two-cell stage. In these larvae (Fig. 10, D), there is present a single gut, of normal size and typical tripartite structure. In one of the two eggs, the cells whose prospective value is gut endoderm have been incorporated into the outer covering of the larva. Although "ectodermal" in the sense of location, these cells continue to differentiate into endoderm, i.e., they do not develop chromatophores and they become ciliated on their outer surface. The self-differentiation of the endoderm in this position indicates, as does the development of exogastrulse (Novikoff, 1938), the complete independence of endodermal and ectodermal differentiation in Sabel- laria vulgaris.
Since the polar lobe, as well as any of the quarter- or half-blasto- meres, does not affect the differentiation of any cell through contact with that cell, it is not possible to consider the polar lobe an "or- ganizer," in the sense of Spemann (cf. Wilson, 1929, pp. 202-205, and Huxley and deBeer, 1934, pp. 171-172). The experiments of Tyler (1930) have demonstrated that when the first cleavage of the Chxtopterus egg is made equal — either through the retraction of the polar lobe into the smaller, AB, blastomere, or through the cleavage furrow dividing both the egg and the polar lobe equally — then the two half-blastomeres are totipotent. In both types of equal cleaving eggs, the AB blastomere receives polar lobe material. If allowed to develop in toto, such eggs produce double monsters of the cruciata type. (At the second cleavage, two very small polar lobes may be formed.) When separated, each half-blastomere produces a fairly normal larva. There is in reality no "AB" cell; each cell behaves like a CD. But there is no evidence that the substance, whose altered distribution
ALEX B. NOVIKOFF
changes the prospective value of the cell from that of an AB to that of a CD cell, is located within the polar lobe. The fact that a double embryo is produced when the polar lobe goes in its entirety into the AB, i.e., that the CD can produce an embryo without the materials of the polar lobe, indicates the complexity of the situation. That the vegetal hemisphere of the molluscan or annelid egg possesses a parti- cular substance at the time of the first cleavage is well established by the work on Ilyanassa (Crampton, 1896), on Dentalium (Wilson, 19046), on Tubifex (Penners, 1924, 1926), and on SabeUaria (Hatt, 1932, and the present paper). In Tiibifex, if the first cleavage of the whole egg or the CD blastomere is made to take place equally instead of unequally (through heat or lack of oxygen) — and the pole-plasms are distributed equally to the two cells — double cruciata monsters are produced (Penners, 1924). A single case was observed in SabeUaria in which the first two blastomeres were of equal size. \Yhen isolated, each of the blastomeres produced polar lobes at the next two cleavages, and each gave rise to a larva possessing an apical tuft. The fact that each cell forms polar lobes and that the two cells are of equal size indicates that each cell probably received materials from the first polar lobe. It would, then, appear that some substance present in the first polar lobe does have the ability to change the course of development of a cell, but that this substance does not act by contact with a cell ; it must become a part of the cell. Normally, only the CD cell develops an apical tuft since the materials of the lobe flow only into that cell. The exceptional production of apical cilia by a CD cell and the appearance of post-trochal bristles in a BC combination, mentioned earlier, may be due to a deviation in either of the first two cleavage furrows, with a resultant unusual allocation of materials. The doubling of the apical tuft in the E-PL2 larva may also be due to an unusual pattern, in which the materials giving rise to apical tufts are separated into differ- ent cells.
In the so-called regulative eggs, it is in many cases possible to alter the course of differentiation of a cell by transplanting the cell to a new position in the developing embryo. By varying the stage at which the operation is performed, the time of determination of a structure may be ascertained. Due to the scarcity of similar experiments on mosaic eggs, relatively little information is available concerning the effect of one part of an embryo on another during the course of develop- ment, or the existence of inducing, or organizing, regions in these eggs. Penners (1926, 1934) destroyed varying numbers of mesodermal and ectodermal teloblasts of Tubifex, at different stages, to test (1) the inter-dependence of ectoderm and mesoderm during development, and
DETERMINATION IN SARELLARIA 233
(2) the inducing capacity of the teloblasts. He found that the ecto- derm and mesoderm show complete independent differentiation, except for a slight influence of the development of one upon the form and upon the rate of development of the other. The fact that following the destruction of the teloblasts the embryos continue to develop normally indicates that the teloblasts are not organizing centers. Horstadius (19376) combined various quartets of blastomeres of the sixteen-cell stage in Cerebratuhis; he found no effect of one layer upon the differentiation of the others. These results are in agreement with those of the present investigation, in which the polar lobe, half-, and quarter- blastomeres are shown to be ineffective in directing the development of Sabellaria.
However, in another egg which was thought to be mosaic, the egg of the ascidian, effects of one cell upon the development of the other have been reported. Tung (1934) found that some factor outside the brain is responsible for the formation of the sense organ, in Ascidiella. Also, Tung found indications that the adhesive organ is induced. A recent paper by Rose (1937) reports that, similarly, in the egg of Styela, the eye spot is induced by the gray macromeres.
SUMMARY
1. Isolation experiments on the egg of Sabellaria vulgaris demon- strate that the formation of the apical tuft in partial larvae is dependent upon the presence of the first polar lobe and the C cell ; that the post- trochal region develops only when the three polar lobes and the ID cell are present; and that apical cilia form only if the A or B cell is included.
2. Form changes in isolated first and second polar lobes are de- scribed. The early changes are synchronous with the cleavages of the ovum, except that all events in the isolated lobe are delayed.
3. The results of the following transplantation experiments are reported: (a) Transplantation of polar lobes, (b) Transplantation of blastomeres to the whole egg. (c) Transplantation of blastomeres to E-PL1. (d) Fusions of half- and quarter-blastomeres. (e) Fusions of two eggs. In all combinations, complete self-differen- tiation of individual blastomeres occurs. Apical tufts develop only when C cells are present, post-trochal bristles are dependent upon the presence of D cells, and apical cilia form only when either the A or B cell is included.
4. The results of this investigation are compared with those from experiments on other mosaic eggs.
It is with pleasure that the writer expresses his gratitude to Pro-
234 ALEX B. NOVIKOFF
fessor L. G. Earth for his untiring assistance and constant encourage- ment throughout the course of this investigation, and to Professor E. B. Wilson for his inspiring interest in the work.
BIBLIOGRAPHY
CRAMPTON, HENRY E., 1896. Experimental studies on gasteropod development. Arch. f. Entw.-mech., 3:1.
DELAGK, YVKS, 1899. Etudes sur la merogonie. Arch, de Zool. exper. el gen. (Ser. Ill), 7: 383.
HATT, PIERRE, 1932. Essais experimentaux sur les localizations germinales dans 1'oeuf d'un Annelide (Sabellaria alveolata L.). Arch. d'Anat. Micros., 28: 81.
HORSTADIUS, SVEN, 1937a. Microdissection, Free-hand Manipulations. In Hand- book of Microscopical Technique, edited by C. E. McClung. Paul B. Hoeber, p. 43.
HORSTADIUS, SVEN, 19376. Experiments on determination in the early development of Cerebratulus lacteus. Bid. Bull., 73: 317.
HUXLEY, JULIAN S., AND G. R. UKBEER, 1934. The Elements of Experimental Embryology. Cambridge. University Press.
MORGAN, THOMAS H., 1933. The formation of the antipolar lobe in Ilyanassa. Jour. Exper. Zool., 64: 433.
MORGAN, THOMAS H., 1935. The rhythmic changes in form of the isolated anti- polar lobe of Ilyanassa. Kiol. Bull., 68: 296.
NOVIKOFF, ALEX B., 1937. Sabellaria vulgaris. In Culture Methods for Inverte brate Animals, edited by Galtsoff et al. Comstock Publishing Co. Inc., p. 187.
NOVIKOFF, ALEX B., 1938. Embryonic determination in the annelid, Sabellaria vulgaris. I. liiol. Bull., 74: 198.
PENNERS, ANDREAS, 1924. Experimentelle Untersuchungen -/.urn Determinations- problem am Keim von Tubifex rivulorum Lam. I. Arch. Mikr. Anat. u. Entw.-mech., 102: 51.
PENNERS, ANDREAS, 1926. Experimentelle Untersuchungen zum Determinations- problem am Keim von Tubifex rivulorum Lam. II. Zeitschr. f. Wiss. Zool., 127: 1.
PENNERS, ANDREAS, 1934. Experimentelle Untersuchungen zum Determinations- problem am Keim von Tubifex rivulorum Lam. III. Zeitschr. f. Wiss. Zool., 145: 220.
ROSE, MERYL S., 1937. The induction of pigment spots in Styela partita (Abstract). Anat. Rec., 70 (Supplement 1): 102.
SCHLEIP, WALDEMAR, 1929. Die Determination der Primitiventwicklung. Leipzig, Akad. Verlags.
TUNG, Ti-Cnow, 1934. Recherches sur les potentialitcs des blastonu-res chez Ascidiella scabra. Arch. d'Anat. Micros., 30: 381.
TYLER, ALBERT, 1930. Experimental production of double embryos in annelids and mollusks. Jour. Expt-.r. Zool., 57: 347.
WILSON, EDMUND B., 1904a. Mosaic development in the annelid egg. Science, 20: 748.
WlLSON, EDMUND B., 19046. Experimental studies on germinal localization. I. The germ-regions in the egg of Dentalium. Jour. Exper. Zool., 1:1.
WILSON, EDMUND B., 1929. The development of egg-fragments in annelids. Arch, f. Entw.-mech., 117: 179.
THE RELATION OF MORTALITY AFTER ENDOMIXIS TO THE PRIOR INTERENDOMICTIC INTERVAL IN PARAMECIUM AURELIA l
BERNICE FRANCES PIERSON
(From the Department of Zoology, The Johns Hopkins University)
Investigators of endomixis in Paramecium have frequently observed that animals often die during or soon after endomixis. According to Erdmann and Woodruff (1916), few Paramecium caudatum survived endomixis. In Paramecium aurelia, Caldwell (1933) found that death occurred 2.85 to 9.40 times more frequently at endomixis than in the middle of the period between endomixes. The present paper is a presentation of the results of a study of one of the factors determining such deaths, namely, the length of the preceding interendomictic interval. As will appear below, when the prior interendomictic inter- val is unusually long, endomixis results in a greater percentage of deaths than when the interval is of the ordinary duration. Moreover, the greater the interendomictic interval, the greater is the mortality resulting from endomixis, until, after very long intervals, endomixis invariably results in death.
In order to investigate this question, it was essential to have avail- able, simultaneously, lines with normal interendomictic intervals and sister lines with unusually long intervals. This was accomplished by employing recently developed methods of inducing endomixis (Sonne- born, 1937) and of obtaining lines with long interendomictic intervals (Sonneborn, 1938). To induce endomixis, the surplus animals from daily isolation lines of cultivation were collected in a small amount of fresh culture medium and kept at 31° C. for a few days until endomixis occurred. To obtain lines with long interendomictic intervals, daily isolation lines which went into endomixis were replaced by sister lines which had not yet gone into endomixis. In this way, lines with long intervals are selected for study while the lines with shorter intervals are eliminated.
Using these techniques, the following experiment was performed. The vegetative descendants of a single endomictic individual were cultivated as 24 daily isolation lines of descent for 165 days. During this time, all lines that went into endomixis were eliminated and re-
1 This work was suggested by Dr. T. M. Sonneborn, to whom I wish to express my sincere appreciation for his helpful advice and assistance.
235
236
BERNICE FRANCES PIERSON
placed by surplus animals from sister lines; so that at all times this group (I, Fig. 1) consisted of 24 lines that had not been in endomixis since the start of the experiment. At five successive intervals of 21 to 31 days, surplus animals from this group were induced to go into
Time from initiol endomixis, in days
|
0 3 |
1 5 |
6 6 |
1 1C |
2 12 |
b 165 T |
|
A' |
H ie>< |
'- 1C |
. ID' |
^- IE |
u |
|
IE> i i |
i— DC |
r- ID |
r- IE |
||
|
i >v II Oe |
: 12 Jar - K |
.6 Jar 25, |
31 feb 21 |
^ 21 Mor 25 |
•. 15 Apr. 24 |
Intervals between successive cndomixes, m days
FIG. 1. Plan of Experiment
The horizontal lines (solid) represent groups of isolation culture lines. Long, solid line " I " represents a group of 24 isolation lines from which all endomictics were eliminated. Long, solid line " II " represents a group of 24 isolation lines, interrupted periodically (at points marked e) for the induction of endomixis in small mass cultures. The short, solid, horizontal lines represent groups of isolation lines carried for 15 generations. The length of the short lines corresponds to the average number of days that the 15 generations lived in all the groups.
The vertical lines (broken) connect the source groups (represented by long, solid, horizontal lines) with their derived groups (represented by short, solid, horizontal lines). In each case, the derived group began with animals in endomixis taken from mass cultures of animals from the source group.
A, B, C, D, and E are the designations given to the derived groups from both I and II. The four derived pairs of groups of isolation lines between which comparisons are made art.- groups In, Iln; Ic, lie; ID, I1|>; IK, UK. Group II was derived from Group I.\ by continuing, without endomixis, 24 of the lines of this group.
e stands for climax of endomixis. This symbol is put in on the day on which endomictics were isolated from the mass cultures.
T indicates the time of termination of the experiment.
Time in days, from the initial endomixis in Group I, is shown on the horizontal axis at the top of the diagram. The number of days between successive endomixes is shown at the bottom of the diagram.
endomixis and the resulting mortality in them was determined. In these groups (IA, In, Ic, ID, and IK, Fig. 1), therefore, tin- mortality after interendomictic intervals of 31, 56, SI, 102, and 125 days, re- spectively, was ascertained. At the same time that each of these
MORTALITY AND ENDOMIXIS IN PARAMECIUM 237
groups (except the first) was being studied, there was examined a control group of lines (hat went through endomixis at the same time, but had instead an interendomictic interval of normal extent (21 to 25 days). The four control groups (I IB, He, I In, and HE) were ob- tained by inducing four successive endomixes at the proper time in descendants of the group I.\. Comparisons of mortality were thus made between the following pairs of groups: IB (previous interval 56 days) with HB (previous interval 25 days); Ic (previous interval 81 days) with He (interval 25 days); ID (interval 102 days) with IID (interval 21 days); and IE (interval 125 days) with II E (interval 23 days).
In this experiment, the animals employed were all descended vegetatively from one which was isolated from a Johns Hopkins stock
TABLE I
Number of fissions from climax of endomixis until death in sixty non-viable lines among one hundred and twenty exendomictic lines followed through until death or the next induction of endomixis.
(In these lines, the maximum interendomictic interval was 71 fissions )
Number of Fissions
until Death Frequency
|
1 |
2 |
|
2 |
3 |
|
3 |
4 |
|
4 |
6 |
|
5 |
13 |
|
6 |
7 |
|
7 |
7 |
|
8 |
6 |
|
9 |
6 |
|
10 |
1 |
|
14 |
1 |
|
27 |
. . . . 1 |
|
29 |
1 |
|
30 |
1 |
|
37. |
1 |
mass culture of Woodruff's long-lived Yale race. The organisms were cultivated throughout according to the methods described by Sonne- born (-1936). Samples of all lines, except those not carried through till the next endomixis, were stained daily to determine the nuclear con- dition.
Relationship of the Interendomictic Interval to the Percentage
of Mortality after Endomixis
The percentage mortality was computed as follows: In each group, each endomictic animal was cultivated as a single daily isolation line until the fifteenth fission after the climax of endomixis. The per-
BERNICE FRANCES PIERSOX
centage of lines that died during this period was the percentage mor- tality for the group. The period of 15 fissions was decided upon as a conservative standard in view of the facts that the process of endomixis itself lasts not more than 9 or 10 fissions and that experience of previous investigators (e.g. Caldwell, 1933) as well as our own showed that deaths rarely occur in the interval from the tenth generation after the climax of one endomixis until the following endomixis. Table I gives i In- frequency of death at various stages of the interendomictic interval in the material here investigated. As appears from the table, deaths occurred but rarely after the tenth fission.
The results of the experiment are summarized in Tables II to IV. Table II gives a general view of the relation of mortality after endo- mixis to the length of the previous interendomictic interval. As appears in the table, the percentage mortality increases as the previous interendomictic interval increases. It rose from 32.2 per cent after
TABLE II
Relation of mortality after endomixis to length of previous interendomictic interval.
|
Previous Interendoiuii i ic Interval in Days 21 |
Percentage of Mortality after Endomixis 32.2 |
Number of Endomictic Lines Observed 90 |
|
23 |
62.5 |
48 |
|
25 |
.... 59.3 . . . |
91 |
|
25 |
57.0 |
79 |
|
31 |
66 6 .... |
48 |
|
56 . |
. . 71.4 . |
49 |
|
81 |
79.2 |
53 |
|
102 |
89 3 . . . . |
84 |
|
125. |
.100.0. . |
.29 |
an interval of 21 days to 100 per cent after an interval of 125 days. The first five intervals, 21, 23, 25, 25, and 31 days are normal inter- endomictic intervals for Woodruff's stock of Paramecium anrelia. The last four intervals of 56, 81, 102, and 125 days are abnormally long intervals for this stock. It will be noted that even in normal inter- endomictic intervals the percentage of mortality was high.
It will be observed, from an examination of Table II, that there may be great differences in mortality after endomixis even when the inter- vals between endomixes are practically the same. Such a difference appears between the percentages of mortality after the 21- and 23-day intervals when the percentages of mortality were 32.2 per cent and 62.5 per cent, respectively. Obviously there are factors oilier than inter- endomictic interval involved in the production of mortality. Knviron- mental factors probably play an important role here. Therefore, it
MORTALITY AND ENDOMIXIS IN PARAMECIUM
239
was essential, for the purposes of this study, to make simultaneous comparisons of groups kept under identical cultural conditions.
The results of such simultaneous comparisons are given in Table III. There the percentages of mortality from endomixis after four abnor- mally long interendomictic intervals are compared with the percentages of mortality from concurrent endomixes after four normal intervals. Thus, when the interendomictic interval was 56 days, there was 71.4 per cent mortality, as compared with 59.3 per cent in the concurrent group with a normal interval of 25 days. When the interval was increased to 81 days, the mortality was between 79.2 and 84.0 per cent (the exact figure depending upon how many of 20 animals that died without dividing — and hence without a determination of whether they had been in endomixis — were in endomixis), as compared with
TABLE III
Relation between length of previous interendomictic interval and percentage
of mortality after endomixis.
|
Groups with Normal |
Groups with Long |
||||
|
Interendomictic Intervals |
Interendomictic Intervals |
||||
|
Previous In- |
Percentage |
Percentage |
Previous In- |
||
|
Name of |
terendomictic |
Mortality |
Mortality |
terendomictic |
Name of |
|
Group |
Interval in |
after |
after |
Interval in |
Group |
|
Days |
Endomixis |
Endomixis |
Days |
||
|
HB |
25 |
59.3 |
71.4 |
56 |
IB |
|
He |
25 |
57.0 |
79.2 |
81 |
Ic |
|
IID |
21 |
32.2 |
89.3 |
102 |
ID |
|
HE |
23 |
62.5 |
100.0 |
125 |
IE |
57.0 per cent in the concurrent group with a normal interval of 25 days. When the interval was still greater, 102 days, the mortality likewise increased to 89.3 per cent as compared with 32.2 per cent in the controls with a normal interval of 21 days. Finally, when the interval reached 125 days, the mortality was 100 per cent, as compared with 62.5 percent in the controls with a normal interval of 23 days. The effects of still greater intervals could not be studied because efforts to induce endo- mixis at intervals of 148, 158, and 161 days all failed; all but a few of the animals in the induction cultures died within 48 hours.
The results of the experiment, as summarized in Tables II and III, show clearly that as the time between two successive endomixes in- creases, the percentage of lines that are unable to survive the second endomixis also increases until eventually no animals can survive.
240
BERXICE FRANCES PIERSOX
Relationship of the Interendomictic Interval to the Number of Generations
between Endomixis and Death
The relationship between the length of the previous interendomictic interval and the number of generations which the non-viable exendo- mictic lines lived after the climax of endomixis is shown by the data in Table IV. When the previous interval was of normal extent, i.e. about 25 days, as was the case in the control groups, the mean number of generations which the lines lived after the climax of endomixis re- mained fairly constant, ranging only from 5.0 to 5.9 generations.
TABLE IV
Length of life in number of generations from climax of endomixis until death, in the non-viable exendomictics, in relation to prior interendomictic interval. (The number dying without fission wa^ not determined in experiments 1 and 2.)
|
Ex- peri- ment |
i troup |
Inter- endo- mictic Inter- val in Days |
Number of Generation-; from Climax of Endomixis until Death |
To- tal Non- viable Exen- domic- tics |
Mean |
|||||||||||||||
|
0 |
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
0 |
10 |
11 |
12 |
13 |
14 |
15 |
|||||
|
1 |
Us |
25 |
? |
1 1 |
2 1 |
s |
5 |
15 |
7 6 7 |
8 |
8 2 |
4 |
1 |
54 |
5.9 |
|||||
|
IB |
56 |
? |
7 |
'> |
5 |
3 |
1 |
35 |
4.8 |
|||||||||||
|
2 |
He |
25 |
? |
1 |
1 |
5 |
6 |
13 |
4 |
2 |
1 |
1 |
2 |
1 |
1 |
45 |
5.8 |
|||
|
Ic |
81 |
? |
7 |
3 |
3 |
6 |
9 |
8 |
4 |
1 |
1 |
42 |
4.4 |
|||||||
|
3 |
HD |
21 |
3 |
1 |
1 |
3 |
3 |
3 |
7 8 |
2 4 |
1 |
1 |
2 |
1 |
29 |
5.0 |
||||
|
ID |
102 |
6 |
1 |
5 3 |
L3 |
18 |
l > |
1 |
4 |
2 |
2 |
75 |
4.4 |
|||||||
|
4 |
HE |
23 |
4 |
4 |
3 |
5 |
1 |
2 |
2 |
1 |
1 |
1 |
1 |
30 |
5.5 |
|||||
|
IE |
125 |
21 |
1 |
2 |
2 |
2 |
1 |
29 |
0.9 |
Among the experimental groups ihr mean number of generations decreased from 4.8 general inns when the interval was 56 days to 0.(J generation when the interval was 125 days.
In contrast to the steady increase in mortality rate with increasing interendomictic intervals, the mean survival period shows no such steady change. When the interendomictic interval was twice as great as normal the survival period decreased but little. Kven when the interendomictic interval was 3 or 4 times as great as normal, the sur- vival period was decreased, on the average, by only about one fission.
MORTALITY AND ENDOMIXIS IN PAKAM 1'XTUM 241
Differences of this magnitude are of doubtful significance because the method of determining when the climax of endomixis occurred involved an uncertainty. No direct observation could be made on the endo- mictic animals removed from the induction cultures to begin the experimental and control groups; but the next day one or more products of their fission were stained and the nuclear condition recorded. On the basis of the number of fragments of the old macro- nucleus, their size and the intensity of stain, and the size, form, and intensity of stain of the new macronucleus or its anlage, an estimate could be made of how many fissions had occurred since the climax of endomixis; but such an estimate may often be in error by one or two fissions. For this reason, little significance is attached to the slight differences among the preceding groups in the mean number of genera- tions that the non-viable lines survived.
In the group with the longest interendomictic interval, however, the difference is so great as to be unquestionably significant. After an interendomictic interval of 125 days, the non-viable exendomictics went through only one-sixth as many fissions as the non-viable exendo- mictics with normal prior interendomictic intervals. Indeed, 72.4 per cent of them failed to divide at all after the climax of endomixis. On the third day without fission they were all stained and found to contain fragments of the old macronucleus, but no new anlage or macronucleus. It might be suggested that about three-fourths of the parent lines had lost their micronuclei before endomixis was induced. In such animals only the destructive phases of endomixis can take place as there is no reserve micronucleus from which a new macronucleus can be formed. Loss of the micronucleus after long omission of endomixis has in fact been observed by others in this laboratory (unpublished).
Discussion
1. In view of the complex and superficially paradoxical relations between endomixis and mortality, it has sometimes been held that endomixis is neither a definite phenomenon nor a normal one, but that it is a pathological response of the organism to adverse conditions, the type of response and its consequences varying with the degree of unfavorableness of the environment. In the present work, effects due to differences of environment were avoided by systematically exchang- ing culture medium between the various groups compared and by restricting comparisons to groups examined at the same time under the same cultural conditions. Thus, the differences in mortality after endomixis were not consequences of environmental differences. Nor were they due to the cumulative action of unfavorable conditions, for :
242 BERXICE FRANCES PIERSON
each case the groups compared had been subjected equally long to the same conditions. The differences in mortality after endomixis must therefore have been due to the difference in the prior interendomictic interval. The fact that mortality is increased after longer intervals shows that the frequent occurrence of endomixis is an advantage to the stock. In view of this, the interpretation of endomixis as pathological is untenable. Similar conclusions were reached by Kimball (1937) from a study of the precise ratios in which sex segregates after endo- mixis in this species.
2. The vit-w that Protozoa are potentially immortal and that natu- ral death does not exist among them became popular during the last quarter of the nineteenth century. It has long been known, however, that while a race as a whole may be potentially immortal, certain mem- bers of the race are doomed, from internal causes, to die. Thus, Jennings (1913) and others showed that conjugation often resulted in unavoidable death. Erdmann and Woodruff (1916), Jennings, Raffel, Lynch, and Sonneborn (1932), Raffel (1932),,Caldwell (1933), Sonne- born and Lynch (1937), and others have shown that endomixis likewise often results in death. \Yoodruft (1917), Sonneborn (1935), and Jennings and Sonneborn (1936) have shown that long omission of endo- mixis ultimately results in death. To these intrinsic causes of death in Paramecium the present paper adds another: The mortality at endo- mixis is directly proportional to the preceding interval without endomixis.
SUMMARY
1. Using an interendomictic interval of 20 to 30 days as a standard, it was shown that intervals approximately two, three, four, and five times this long resulted in progressive increases in mortality after endomixis until 100 per cent mortality occurred.
2. At intervals greater than this, the animals died before endomixis could be induced under conditions favorable for its induction.
3. The mean number of generations which non-viable lines sur- vived the climax of endomixis was 5.0 to 5.9 generations when the pre- vious interval was of normal extent. After an interval of 125 days, survival dropped greatly to a mean of 0.9 generation.
4. The results are shown to disagree with current interpretations of mortality at endomixis based on the concept of endomixis as a patho- logical process.
5. The results show that unusually long interendomictic intervals are, like other previously known conditions, a cause of " natural death " in Protozoa.
MORTALITY AND ENDOMIXIS IN PARAMECIUM 243
LITERATURE CITED
CALDWKLL, L., 1933. The production of inherited diversities at enclomixis in
Paramecium aurelia. Jour. Exper. Zool., 66: 371. ERDMANN, R., AND L. L. WOODRUFF, 1916. The periodic reorganization process in
Parannpcium caudatum. Jour. Expcr. Zool., 20: 59. JENNINGS, H. S., 1913. The effect of conjugation in Paramecium. Jour. Exper.
Zool., 14: 279. JENNINGS, H. S., DANIEL RAFFEL, RUTH STOCKING LYNCH, AND T. M. SONNEBORN,
1932. The diverse biotypes produced by conjugation within a clone of
Paramecium aurelia. Jour. Exper. Zool., 62: 363. JENNINGS, H. S., AND T. M. SONNEBORN, 1936. Relation of endomixis to vitality in
Paramecium aurelia. Comptes rendus du XIIC Congres International de
Zoologie (Lisbonne 1935), pp. 416-420. KIMBALL, R. F., 1937. The inheritance of sex at endomixis in Paramecium aurelia.
Proc. Nat. Acad. Sri., 23: 469. RAFFEL, DANIEL, 1932. The occurrence of gene mutations in Paramecium aurelia.
Jour. Exper. Zool., 63: 371. SONNEBORN, T. M., 1935. The relation of endomixis to vitality in Paramecium
aurelia. Anal. Rcc., 64: (Supplement No. 1) 103. SONNEBORN, T. M., 1936. Factors determining conjugation in Paramecium aurelia,
I. The cyclical factor: the recency of nuclear reorganization. Genetics, 21:
503. SONNEBORX, T. M., 1937. Induction of endomixis in Paramecium aurelia. Biol.
Bull., 72: 196. SONNEBORN, T. M., 1938. The delayed occurrence and total omission of endomixis in
selected lines of Paramecium aurelia. Biol. Bull., 74: 76. SONNEBORN, T. M., AND R. S. LYNCH, 1937. Factors determining conjugation in
Paramecium aurelia. III. A genetic factor: the origin at endomixis of
genetic diversities. Genetics, 22: 284. WOODRUFF, L. L., 1917. The influence of general environmental conditions on the
periodicity of endomixis in Paramecium aurelia. Biol. Bull., 33: 437.
THE EFFP:CT OF SHORTFR THAN NORMAL IXTER-
ENDOMICTIC INTERVALS ON MORTALITY AFTER
ENDOMIXIS IX PARAMECIUM AURELIA
JULIUS GELBER1 (From the Department of Zoology, The Johns Hopkins University)
In the preceding paper, Pierson (1938) has shown for Parameciiim aurclia thai mortality after endomixis is directly proportional to the length of the interendomictic interval, when the latter is longer than normal. The present paper examines the same question when the interendomictic intervals are shorter than normal.
The same race of Paramecium aurelia is examined in both studies. All individuals in the present study were descended without endomixis from one taken from a stock mass culture of this race on February 22, 1936. The methods of culture and of induction of endomixis were similar to those employed by Pierson.
The experiment was performed as indicated in Fig. 1. From a single endomictic individual, 24 daily isolation lines of cultivation were followed for 28 days (Group I, Fig. 1). During this time, samples of each line stained daily showed that no endomixis occurred. On the eighth and sixteenth days after the initial endomixis, the surplus animals from the isolation lines were collected in a mass culture and placed at 31° C. In each case, a high percentage of individuals went into endomixis after 48 hours in these conditions. From each of these, a group of endomictic individuals was isolated and cultivated (Groups II and III, Fig. 1). On the twenty-sixth day after the initial endo- mi\K iliree mass cultures were set up, each consisting of the surplus animals from one of the three groups under cultivation. These cul- turo were placed at 31° C. and all contained numerous endomictic individuals two days later. From each of the three cultures a group of endomictic individuals was isolated and cultivated. These three groups of animals \\ere thus all in endomixis at the same time, but they dittncd in the interval since the last preceding endomixis: in one, the interval was 10 days; in the second, 18 days; and in the third, 28 days.
The three groups will be designated the 10, 18, and 28-day groups, in reference to their prior interendomictic intervals. The 10-day
1 The author is grateful to Dr. T. M. Sonneborn who suggested the problem and offered his advice throughout the experiment, and to Bernice Pierson and Nathaniel Finkelstein for their kind assistance.
244
MORTALITY AND ENDOMIXIS IN PARAMECIUM 245
group was begun with 192 individuals from the culture in which endo- mixis had been induced; but only 51 of these were in endomixis, as determined by staining products of their fissions on the next day. From each of these 51 individuals, a single daily isolation line was cultivated until it died or until 15 successive fissions had taken place. Of the 51 lines, 48 lived through the 15 fissions and three died after 5 to 12 fissions, giving a mortality rate of 5.9 per cent.
The 18-day group was begun with 144 individuals from the induc- tion culture; but only 94 of these were in endomixis, as subsequently determined. Of these 94 endomictic individuals, 74 lived through the following 15 fissions, and 20 died after 1 to 7 fissions, giving a mortality rate of 21.3 per cent.
Time in Days
Percentage
Mortality
41.8 5.9
21.3
i v i w ** Y
FIG. 1. Plan of Experiment
The three horizontal lines, I, II, and III, represent groups of isolation culture lines carried without endomixis for 28, 18, and 10 days, respectively. The vertical lines connect the source group (horizontal line I) with the two derived groups (hori- zontal lines II and III). In each case the derived group began with animals in endo- mixis taken from mass culture of animals from the source group.
e stands for endomixis.
The upper line shows the time in days since the initial endomixis in Group I.
The percentage mortality following endomixis in each group is shown to the right of the line representing that group.
The 28-day group was begun with 144 individuals from the induc- tion culture. Of these, 110 were shown by subsequent staining of their descendants to have been in endomixis. These 110 endomictic individuals were cultivated in the same way as were those of the other two groups. In this group, 64 lived through the 15 fission period of observation and 46 died after 0 to 13 fissions, giving a mortality rate of 41.8 per cent.
Thus, the group with a normal interendomictic interval of 28 days suffered a mortality rate of 41.8 per cent after endomixis, as compared with 21.3 per cent and 5.9 per cent mortality following abnormally short interendomictic intervals of 18 and 10 days, respectively.
The results therefore extend those of Pierson to include abnormally
|
0 10 18 28 I ...,,.... 1 ....... . «_l |
(28 day [Group (lO day (Group fl8 day \G r o u p |
|
• n |
m. 0 |
|
1 C |
246 JULIUS GELBER
short as well as abnormally long interendomictic intervals. Through- out the entire range of intervals investigated, the mortality after endomixis is directly proportional to the extent of the preceding inter- endomictic interval.
LITERATURE CITED
PIERSON, BERNICE F., 1938. The relation of mortality after endomixis to the prior interendomictic interval in Paramecium aurelia. Biol. Bull., 74: 235.
DIPLOIDS FROM UNFERTILIZED EGGS IN HABROBRACON
KATHRVN G. SPEICHER AND H. R. SPEICHER1
(From the Department of Zoology, University of Maine, and the Marine Biological Laboratory, Woods Hole, Mass.)
FEMALES
The occurrence of impaternate females (females from unfertilized eggs) in the parasitic wasp Ilabrobracon juglandis (Ashmead) has previously been reported (Speicher, 1934). Such females occur sporadically from various virgins, and regularly constitute about 1 per cent of the F2 population from Fi virgins produced by outcrossing females from tapering or reverted tapering stocks. The hypothesis was offered that these diploid impaternate females might be produced by the failure of the second maturation division in the unfertilized egg and, carrying two chromatids from a single tetrad, would provide material for studying the mechanics of crossing-over.
When FI virgins were heterozygous for recessive factors, ¥2 im- paternate females occurred in the ratio of one homozygous dominant, to two heterozygotes, to one homozygous recessive for each locus involved. This 1:2:1 ratio was at variance with results obtained in other organisms where more than one strand may be recovered from a single tetrad, notably Drosophila (Anderson, 1925) and Neurospora (Lindegren, 1933). Here the first maturation division is reductional at the spindle fiber and in dyads from that division the homozygosis of any locus depends upon the amount of crossing-over between it and the fiber, and hence is a function of its location along the chromosome. In Ilabrobracon, except for the locus of the sex-linked factor fused (Whiting and Speicher, 1935), the amount of homozygosis was con- sistently 50 per cent for all loci tested, even including those of two recessives known to be linked and separated by a distance of ten units.
It was realized from the beginning of the work that other hypoth- eses could be advanced to explain the formation of impaternate females. And it \vas apparent that the 1:2:1 ratio obtained would be expected for all loci if the two homologous chromosome strands of an impaternate female came from two tetrads, independent and com-
1 The authors are indebted to Professor P. W. Whiting of the University of Pennsylvania for the use of special microscopic equipment which was supplied in part by a grant to him from the Elizabeth Thompson Science Fund.
247
248
KATIIKVN G. SPEICHER AND B. R. SPEICHER
pletely reduced, rather than from a single tetrad which underwent only partial reduction. As an investigation of oogenesis in Ilabrobracon had already been begun (Speicher, 1936), it seemed best to suspend further genetic studies until the method of formation of impaternate females could be determined cytologically. Those findings are here reported for the first time.
Since impaternate females had been shown to be genetically diploid, whereas their sibs are haploid, a cytological analysis of their formation seemed practical. According to past work on other forms at least three basic hypotheses could be considered. First, one of the two maturation divisions of an unfertilized egg could be suppressed, as was observed by Silvestri (1908) in the wasp Prospalta. This would leave a cleavage nucleus containing the diploid number of chromatids, which would restore the diploid number of chromosomes if the homo- logues separated. Second, fusion of two reduced egg nuclei present
TABLE I
Progeny from No. 25/reta virgin females.
|
Fi virgin 9 9 |
i |
2* |
3 |
4 |
5 |
6 |
7* |
8 |
9 |
10 |
11 |
|
Sons . ... |
83 |
0 |
01 |
26 |
OS |
Q7 |
s |
67 |
9Q |
11 * |
6* |
|
Daughters |
1 |
2 |
0 |
0 |
5 |
0 |
0 |
0 |
o |
0 |
7 |
|
Eggs collected |
4Q |
3 |
50 |
S6 |
46 |
61 |
11 |
7Q |
48 |
S6 |
S] |
* Died before completion of experiment.
in a binucleate egg also would restore the diploid condition. Third, the egg might originate as a tetraploid, undergo reduction and thus become diploid. A fourth hypothesis, the doubling of chromosomes in the haploid egg during cleavage, is eliminated because tests have shown that impaternate females may be genetically heterozygous.
The first two theories were tested together. Over 300 eggs were collected from virgin females produced by crossing reverted tapering and stock 25. They were fixed at first cleavage prophase and stained by the Feulgen whole-mount method. The same females produced collectively 15 impaternate daughters among 724 sons, over 2 per cent of the total. Among the eggs studied approximately the same pro- portion would be expected to show cytological differences if either theory were correct. Suppression of a maturation division would re- sult in a decrease in the normal number of polar nuclei formed at the egg margin. A binucleate egg would be expected to show two groups
DIPLO1DS, UNKKKTIUZED EGGS OK HABROBRACON 249
of polar nuclei, totaling twice the normal number. An examination of all eggs revealed none showing either of these two differences.
Treatment of the third hypothesis requires chromosome counts during maturation; accordingly it was necessary to obtain a new lot of eggs fixed at an earlier stage than the above. It had been noticed previously that some virgin females produced impaternate daughters in small groups while others in the same experiment produced no daughters whatever. Inclusion in the data of the offspring from the latter virgins lowers the percentage of impaternate females among total offspring. It was therefore possible to raise the expected percentage of exceptional eggs by selecting eggs only from virgins known to be producing impaternate daughters. This was accom- plished as follows. Eleven females from a cross of reverted tapering by stock 25 were placed with host caterpillars. Eggs laid upon the caterpillars during seven consecutive days were fixed at first anaphase and temporarily stored in alcohol. Eggs laid at night over the same period were allowed to develop, in order to indicate which of the eleven females were thelytokous. Results are shown in Table I.
;•
Kig. 1. Fig. 2.
FIG. 1. First anaphase of normal egg. X 3,000. FIG. 2. First anaphase of tetraploid egg. X 3,000.
Seven females, producing a total of 428 sons, had no daughters, while the remaining four produced 241 sons and 15 daughters. The 149 eggs collected from the latter four females were then stained by the Feulgen whole-mount technique. One hundred and ten of these were in condition to study; the remainder were either lost in handling or were collapsed. Ninety-eight eggs were unquestionably diploid, seven were unquestionably tetraploid and five more were questionably tetraploid. The clear cases of tetraploidy showed twenty chromo- somes, presumably bivalents although individual chromatids have never been observed in Habrobracon oogenesis due probably to their small size, moving to one or to each pole. Normal diploid eggs show only ten chromosomes going to each pole, Figs. 1 and 2.
These cytological data, and the fact that impaternate females come in groups from certain mothers suggest the probability that production of tetraploid eggs, as developed from patches of tetraploid
250
KATHRYX G. S PHI CHICK AND B. R. S PIC 1C HICK
ovarian tissue, is responsible for the appearance of cliploid impaternate females among haploid brothers.
M.M.KS
1 >iploid males of biparenial origin have been reported repeatedly in Ilabrobracon. Since diploid females are produced by virgins as a result of tetraploidy in egg cells, it seems plausible that diploid im- paternate males may be produced in the same way. According to Whiting's scheme of sex-determination in Ilabrobracon (AYhiting, 1933) those eggs which were diploid after reduction and contained sex chromosomes A" and Y would produce impaternate females, while those which contained chromosomes XX or YY would produce diploid impaternate males. If distribution of chromatids is random the number of diploid females and males produced should be equal.
TABLE II
I >.itu from tests for diploid impaternate males.
|
K, Q Q |
I- li.tpl'nd c?d\ |
I |
"•j impaternate 9 |
? |
|
.1 or (i |
.1,1 |
Aa |
aa |
|
|
o'/od |
62" |
6 |
10 |
5 |
|
Le/le |
6070 |
1 |
7 |
6 |
|
St/st |
4408 |
4 |
2 |
0 |
Three experiments were set up in an effort to produce diploid impaternate males that could be distinguished genetically from their haploid brothers. Since the highest percentages of impaternate FZ diploids had previously resulted from outcrosses of tapering and reverted tapering females, tapering was again used as the maternal stock in one experiment. In the other two it was necessary to intro- duce recessive factors into the cross through the females. In order to insure the occurrence of thelytoky in these cases the recessives were repeatedly bred up to reverted tapering, and a stock related to it but having the desired genetic constitution was then derived and supplied the females for the parental crosses. In each experiment FI virgins were produced carrying both of two allels that give rise to a distinct phenol \ pe when they are heterozygous.
Results are given in Table II where the first column shows the pairs of allels u-e<l to in, ike the heterozygotes recognizable phenotypically, the second column gives the number of normal haploid males in the FZ, and the third column gives the classes of impaternate females. Theo-
DIPLOIDS, UNFKRTIU/KI) KCOS OK I IABROBRACON 251
retically diploid impaternate males should have occurred in the same numbers and in the same genetic ratio (1 homozygous dominant, AA : 2 heterozygotes, Aa : 1 homozygous recessive, aa) as did the impaternate females, although among the males only the hetero- zygotes (Aa) would have been distinguishable from their normal haploid brothers.
No diploid males were found and the experiments are published at this time because the stocks with which they might be repeated are no longer extant and the work is finished unless or until the thelytokous tendency reappears in Ilabrobracon. However, results are not con- sidered entirely convincing, nor is the evidence against the occurrence of diploid impaternate males considered conclusive. The following paragraphs tell for each experiment why the tests were not as satis- factory as had been expected.
oi/od, — Wasps heterozygous for the allelomorphic eye colors ivory (0*) and dahlia (o'1} had previously been distinguished phenotypically by a lighter color than the dark red of homozygous dahlia (Torvik, 1931). In this experiment light dahlia was recognizable in the FZ females, where it was checked by breeding tests. In the Fo males, however, the color varied widely, grading from dark to light. Forty males were recorded as having eyes as light as their ten heterozygous sisters.
FI virgins in the eye color experiment were also heterozygous for the sex-linked factor fused (fu, antennae and tarsi). Diploid impater- nate daughters would be heterozygous for the sex region (XY), and also for fused except where cross-overs had occurred betewen fused and the sex region (10 per cent of the cases). Of the 21 impaternate fe- males produced none was homozygous fused, and wherever it was possible to make breeding tests fused males occurred among their progeny, so that actually no cross-over cases were found. Diploid impaternate males, on the other hand, would be either XX or YY, and homozygous for/w or Fu except where cross-overs had occurred. Thus almost half the diploid males would carry two fused genes. C. H. Bostian had previously noted that the presence of two fused genes in diploid biparental males frequently made the antennae shorter than those of fused haploid males where only one fused gene occurs. It was hoped that here extreme shortness of antennae might serve as an additional tag for diploidy. However, antennal length proved to be too variable to be of any service, grading gradually from long to short.
Le/le. — In heterozygotes for the gene lemon (le, body color) pigmentation is normal except for the basal segments of the antennae which are yellow instead of dark brown. The dominant factor Min-
KATHRYX G. SPEICHER AND B. R. SPEICHER
nesota yellow (My, base of antenna1) produces exactly the same pheno- type at high temperatures. It is homozygous in reverted tapering stock, and an effort was made to breed it out before this experiment was begun, but its presence was simply hidden by temperature changes. In the F2 both My and le segregated. Of the impaternate females with normal body color it was impossible to tell which carried lemon. All but one were recorded as having yellow antennal bases due to the presence of one or the other or both of the mutant genes. If diploid males heterozygous for lemon were present, it was impossible to dis- tinguish them from their haploid Minnesota yellow brothers.
St/st. — The experiment using the stumpy gene (st, legs) was aban- doned because the number of F2 impaternate females was so small that the expectancy for heterozygous diploid males fell so low as to make the search unprofitable, in spite of the fact that this was the one case where parental females had been of pure tapering stock.
SUMMARY
1. Cytological evidence shows that unfertilized eggs which give rise to females are tetraploid before reduction and therefore diploid after reduction.
2. No diploid males from unfertilized eggs were found among 16,700 males examined.
LITERATURE CITED
ANDERSON, E. G., 1925. Crossing over in a case of attached X chromosomes in
Drosophila melanogaster. Genetics, 10: 403. LINDEGR.EN, CARL C., 1933. The genetics of Neurospora — III. Pure bred slocks
and crossing-over in N. crassa. Bull. Torrey Botanical Club, 60: 133. SILVESTRI, F., 1908. Apunti sulla Prospalta Herlesei, How. e specialmente sui
primi stati del suo sviluppo. Boll. Lab. Zool. R. Sc. Agr. Portici, 3: 22. SPKICHEK, B. R., 1936. Oogenesis, fertilization and early cleavage in Habrobracon.
Jour. Morph., 59: 401. SPKICHER, KATHRVN GILMORE, 1934. Impaternate females in Habrobracon. Biol.
Bull., 67: 277. TORVIK, MAGNHILU M., 1931. Genetic evidence for diploidism of biparental males
in Habrobracon. Biol. Bull., 61: 139. WHITING, P. W., 1933. Selective fertilization and sex-determination in Hymen-
optera. Science, 78: 537. WHITING, P. W. AND KATHRYN GILMORE SPEICHER, 1935. Impaternate daughters
of females heterozygous for a sex-linked gene in Habrobracon. Am. Nat.,
69: 82.
INACTIVATION OF SPERM BY X-RADIATION IN
HABROBRACON
JANE MAXWELL
(From the Marine Biological Laboratory, Woods Hole, Mass.)
Sterility of the male may be due to lack of sperm, to defective or inactivated sperm unable to fertilize the eggs or to sperm carrying dominant lethals which kill the fertilized eggs.
The wasp Ilabrobracon is convenient material for the study of these types of male sterility. Normally males develop from unfertilized eggs by haploid parthenogenesis, females from fertilized eggs. Ab- sence or inactivation of sperm would result in the production of no daughters and of sons equal in number to those of unmated females. If all the sperm contain dominant lethals no daughters are produced and the fecundity of the mated females is correspondingly reduced. Consequently analysis of differences in biparental ratio and in fecundity indicates the type of male sterility which has been operative.
P. W. Whiting (1937) demonstrated that treatment of sperm with X-ray dosages of 20,000, 40,000 and 75,000 r units produces at least one dominant lethal in every sperm cell. No daughters occur in the progeny and the average number of males produced per day does not equal that to be expected from virgin females, indicating that many eggs are fertilized and die. Sperm treated with 75,000 r units fertilized almost as many eggs as untreated sperm ; therefore the treatment ap- parently did not cause inactivation. However, a slight increase in the average males per day from mates of males treated with 75,000 r units as compared with those from mates of untreated controls sug- gested the possibility that spermatogenesis might to some extent be stopped and sperm supply decreased.
An experiment carried out with the X-ray equipment at the Marine Biological Laboratory, Woods Hole, Mass., July-August, 1937, \vas designed to test this hypothesis (Table I). One group of wild-type males (stock 32) was treated with 41,000 r units, another group with 142,000 to 143,000 r units. A third group was untreated. Each male was mated on each successive day to a different female wrhich was
1 The author is indebted to the University of Pennsylvania for furnishing labora- tory space at the Marine Biological Laboratory during the summer of 1937, and to the Committee on Effects of Radiation on Living Organisms (National Research Council) for technical assistance furnished from a grant to Dr. P. W. Whiting.
253
254
JANE MAXWELL
segregated for breeding. I nmated females of the same stock were segregated each day as controls. The total number of days of the egg- laying periods of all females of any one group mated on the same day is the basis of the averages of offspring per day.
It was expected that successive matings would exhaust the sperm of the treated males provided that no spermatogenesis took place after The treatment. Such exhaustion does not occur since the last matings of these males produce average males per day equivalent to those from mates of untreated males and lower than the average for the offspring of unmated females.
TAIJLE I
Offspring from orange-eyed females (stock ll-o) mated with untreated or with x-rayed wild-type males (stock 32) or bred unmated.
|
Day of mating |
Offspring from untreated males |
Offspring from 41.000 r males |
Offspring from 142.000 and 143,000 r males |
||||||||
|
I). IV- |
+ 99 |
Per day |
od"d" |
Per day |
I >.i\ |
od" d" |
Per day |
Day- |
od" o" |
Per day |
|
|
1st |
152 136 145 168 113 93 117 100 59 |
524 671 337 487 303 185 302 258 188 |
3.44 4.93 2.32 2.89 2.68 1.99 2.58 2.58 3.18 |
247 203 116 174 103 62 59 87 82 |
1.62 1.49 .80 L.03 .91 .67 .50 .87 1.38 |
139 129 120 137 120 128 60 128 116 |
184 257 245 223 128 112 44 194 128 |
1.32 1.99 2.04 1.62 1.06 .88 .73 1.51 1.10 |
219 208 120 89 41 |
464 368 285 176 71 |
2.11 1.76 2.37 1.97 1.73 |
|
?nd |
|||||||||||
|
3rd |
|||||||||||
|
4th |
|||||||||||
|
5th |
|||||||||||
|
6th |
|||||||||||
|
7th |
|||||||||||
|
8th |
|||||||||||
|
9th and 10th |
|||||||||||
|
Totals from mated females |
1083 |
3255 |
3.01 |
1133 |
1.05 |
1077 |
1515 |
1.41 |
677 |
1364 |
2.01 |
Total from unmated females. . . .Days, 344. Males, 1073. Males per day, 3.12.
In the total averages of males per day for all the fraternities from treated, evidence is found of failure of sperm to fertilize and thus kill some of the eggs. The total average of males per day increases with increasing dosage and for 142,000 to 143,000 r units is intermediate (2.01) between the mated controls (1.05) and the unmated controls (3.12).
Inactivation of sperm following higher dosages rather than partial exhaustion of sperm supply due to partial inhibition of spermato- genesis is indicated by the fact that there is no increase in males per day from the later matings. It is probable that even much weaker dosages than those here used completely stop spermatogenesis. Thus
INACTIVATION OF SPERM BY X-RADIATION 255
we have partial male sterility due to dominant lethals at relatively weak dosages, complete male sterility clue to dominant lethals at stronger dosages, and partial sperm inactivation at very high dosages. The average offspring per day (males) from unmated females (3.12) is less than the average offspring per day (females and males) from females mated to untreated males (4.05). This is attributed to re- cessive lethals segregating in certain fraternities summarized together in Table I. In other experiments viability of males and females has been about equal.
LITERATURE CITED
WHITING, P. W., 1937. Habrobracon as a means of testing the effectiveness of physical agents in causing mutations. Proc. Penna. Acad. Sci., 11: 50.
THE SEX RATIO IN MELITTOBIA CHALYBII ASHMEAD,
GAMETOGENESIS AND CLEAVAGE IN FEMALES AND
IN HAPLOID MALES (HYMENOPTERA:
CHALCIDOIDEA)
RrnOLF G. SCHMIEDER (From the Zoological Laboratory, University of Pennsylvania)
INTRODUCTION
In the chalcid-fly, Melittobia, females develop only from fertilized eggs, males from unfertilized eggs. The total number of males pro- duced by any individual female is approximately only 3 per cent of the total number of offspring — an unusual condition in view of the fact that these males are indispensable for the continued existence of the species. Not only is the proportion of males very low, but even in virgin females, the number of eggs which develop parthenogenetically into males is not, it is estimated, much over 3 per cent of the total number of eggs laid. The remainder of the eggs die and shrivel up.
When an examination revealed that the males in this species are haploid and that we were not dealing with the common type of faculta- tive parthenogenesis which involves a restoration of diploidy in a limited number of azygous embryos, the question that naturally pre- sented itself was: \Yhy does only such a small percentage of eggs possess parthenogenetic potentialities? As a step in the investigation of this problem cytological observations on gametogenesis, fertilization and cleavage are here recorded.
LIFE HISTORY
Melittobias have been much studied both in America and in Europe. Interest in them has centered about their life-history, their polyphagous habit and the striking structural antigeny they exhibit . More recently (Schmieder, 1933), polymorphism and the factors responsible for this phenomenon have been described in females of M. ckalybii.
These parasites are most frequently found within the cocoons of certain wasps, especially Sceliphron, Chalybion and Trypoxylou. In addition to .17. chalybii, taken in southern New Jersey, two other species, as yet unidentified, have been studied, one from New Jersey, the other from numerous localities in New Jersey, Pennsylvania, Ohio and Indiana. None of these three species when crossed with one an-
256
SEX RATIO IN MELITTOBIA CHALYBII 257
other produced any female offspring but: only matroclinous males. The observations in this paper are all based on M. chalybii material. The other species were, however, also studied and were found to be, cytologically, like chalybii.
In the rearing of these insects, one mated female is placed on a host which has been removed from its cocoon. Feeding and oviposition soon commence and the first 12 to 30 female offspring that emerge oviposit on the same host. Finally, the remainder of the first and the entire second generation (altogether some 500 to 800 offspring) emerge. From 2-5 per cent, usually about 3 per cent, of these are males. Al- though the proportion of males is small and although their number is still further reduced by the mortal combats which occur among them, practically all the females are eventually fertilized. Indeed, the sex ratio here seems to be an adaptive feature which conserves the food supply of the species for the almost exclusive use of the female sex, the sex which alone serves the dispersal of the species. This method of rearing probably approximates what occurs in nature, except that in nature occasional cross-breeding between different lines no doubt may occur while in the laboratory inbreeding is the rule. The stock now on hand is the seventh successive inbred culture. The number of the generation is not known because of the intermingling of two successive generations in each culture. Of one of the other species, with a shorter life history, the twenty-sixth inbred generation is at hand.
Unmated females, segregated during the pupal stage, either may not oviposit at all or may lay only a very few eggs, perhaps from 1 to 5. Of these eggs one or two may develop into males, or all the eggs may die. In an exceptional instance 30 males were obtained from 2 virgins, but the record states that many dried up eggs were also found in this culture.
While virgin females are disinclined to oviposit, such females, after being mated with a male of a different species, generally oviposit as freely as do normally mated individuals. Apparently the act of mating or the presence of sperm in the seminal receptacle acts as a stimulus to oviposition. In no instance, however, did any females develop from such eggs, but only matroclinous males. In mating males of another species to chalybii it is necessary to employ the short-winged poly- morphic form of chalybii whose copulatory reaction time corresponds more closely to that of the foreign males. This short-winged form, however, deposits eggs in very rapid succession and, even when nor- mally mated, an unusually large proportion of these eggs fails to develop.
In the other two species studied no such difficulty exists and in them
RUDOLF G. SCHMIEDER
it is evident, as in chalybii, that only a small proportion, possibly again not more than 3 per cent, develops parthenogenetically into males while all the other eggs die. In one instance 3 females produced over 200 eggs by the sixth day. They were allowed to continue to oviposit until the tenth day but this culture eventually yielded a total of only 20 males. These same females transferred successively to two other hosts, upon which oviposition continued at almost the initial rate, produced on the second host 16 males and on the third host 16 males. The same procedure when carried out with two other females yielded 12 males from the first, 6 from the second and 3 from the third host. No female offspring appeared in any of these cultures.
It is therefore evident that both virgins and females mated to males of another species may produce about the same number of male offspring as do normally mated individuals. Those of their eggs which fail to develop, one is tempted to assume, are those which would have yielded female offspring, had their mothers been mated in the normal way. Without a more adequate statistical analysis, it is not possible to make more definite statements at the present time, especially in view of the high egg mortality often observed even in normal cultures.
MATERIALS AND METHODS
The chalybii material examined consists of: 50 pairs of testes of larvae and pup;e, which were dissected out on glass slips and fixed as smear preparations with Carnoy-Lebrun's and Bouin's fluids; 25 pairs of testes, dissected out in Belar's solution, fixed mainly in Benda's, Bouin's and Petrunkevitch's fluids and sectioned at 4 to 6 ^ in paraffin ; 28 pairs of ovaries of pupa* and imagines, dissected out and fixed directly on glass slips with Carnoy-Lebrun's or Bouin's solution; 220 eggs of normally mated females and 154 eggs of females mated with males of another species, all fixed in Carnoy-Lebrun's medium.
The sectioned material was stained with Heidenhain's haematoxylin. The smears of testes and of ovaries were stained with Mayer's hrema- lum, containing 2 per cent of acetic acid, and by the Feulgen method. For the staining of entire eggs the Feulgen met hod only was entirely satisfactory. All material was mounted in gum damar.
OBSERVATIONS ON MALES
Somatic Tissues
In the cells of the nervous system and of the imaginal discs of pupa? and prepupa- of males there were invariably 5 chromosomes, the hap- loid number (Figs. 1 and 2). This same number was again found in the unfertili/ed eggs in all stages from early cleavage until hatching (Fig. 13).
SEX RATIO IN MKLITTOBIA CHALYBII
Testis
Development of spermatogonial cysts and multiplication of sperma- togonia occur during larval life and only during that period are gonial division stages in evidence. In very young pup;e, in which not even the eye spots as yet show any traces of pigmentation, about one-third of the testis is already occupied by spermatids and mature sperm. Cysts containing the youngest spermatocytes (Fig. 3), arranged in the form of rosettes, occupy the anterior end of the testis.
The middle region is taken up largely by spermatocytes in the growth stage and by division figures. During the growth period the nucleus contains a prominent karyosphere and a lightly staining net- work (Fig. 4). This karyosphere gives way to the fine spireme of the first spermatocyte prophase. As the spireme finally becomes a very open one, the five chromosomes present can easily be distinguished in every cell (Fig. 5). In metaphase plates of spermatocyte divisions (Figs. 6 and 7), it is noted that all five chromosomes are atelomitic. The first division, most clearly recognizable in anaphase (Fig. 8), is abortive and yields only one second spermatocyte. In Fig. 8 and in the figures of the second spermatocyte division (Figs. 9 and 10), only the spindles are shown, which are always clearly distinct from the poorly fixing zone of cytoplasm which surrounds them. The two spermatids formed by division of each second spermatocyte both transform into sperm cells.
OBSERVATIONS ON FEMALES
Somatic Tissues
The cells of the nervous system and of the imaginal discs of female prepupae are always diploid, with 10 chromosomes. The same is true of the follicular cells of the ovary and of the cells of embryos developing from fertilized eggs (Figs. 21 and 34).
Ovary
At the cephalic end of each of the 6 to 8 ovarioles comprising each ovary there is a region in which the oogonia and the future follicle and nurse cells are indistinguishable (Fig. 16). Immediately caudad of this limited region the young oocytes become recognizable by the synaptic stages which are observable in them at a time when these cells are hardly larger than the neighboring presumptive nurse and follicle cells (Fig. 16). In the post-synaptic spireme (Fig. 22), 5 pachytene threads are always clearly discernible. Following synapsis, the oocyte undergoes a steady, continuous growth until maturity,
260 RUDOLF G. SCHMIEDER
while the- follicle- cells and the nurse cells assume the arrangement characteristic of meroistic ovarioles (Figs. 17-20). During this period of growth the chromosomes become more and more deconcentrated and diffuse (Figs. 23-25). In the course of this change they become at first less sharply outlined and open up, the opening up process begin- ning near their ends, in a manner suggesting a longitudinal splitting. Finally large open figures are formed which are most frequently in the form of loose rings. Although this ring-like organization is not always determinable for each of the chromatic elements due to their large size, haziness and frequent discontinuity, yet it is evident that the chromosomes never entirely disappear during the growth period of the
egg-
Toward the end of the growth period the diffuse chromosomes of the germinal vesicle condense and contract into small chromatic bodies which are perhaps suggestive of the precociously formed abortive maturation spindle often described in the literature. (This stage will require additional study.) Finally, when the egg is ready to leave the miduct, the nucleus is represented by a single small karyosphere
H \PLANATIOX OF Fl(,URES
Figures 3, 4, 8, 9 and 10 are from material fixed with Benda's fluid and stained with Heidenhain's haematoxylin.
Figure 6 is from material fixed with Benin's fluid (B 15) and stained with Heidenhain's haematoxylin.
All other figures, on both plates, are from material fixed with Carnoy-Lebrun. Figures 16, 17, 21-24 are from slides stained with acid haemalum; the remainder from slides stained by the Feulgen method.
Magnification: All figures of individual cells or nuclei (Figs. 1-11, 13, 14, 21-26, 28, 30, 32, 34), X 5,000.
Figures of eggs (Figs. 12, 15, 27, 29, 31, 33, 35), X 500.
Figures of the ovariolc 'Figs. 16-20), X 400.
PL ATI: I
FK;. 1. Somatic cell of larva, male; late prophase.
Fl<.. 2. Somatic cell <>i l,n\,i, male; metaphase.
l-ii.. .^. Young spermatocyte, from a rosette.
I ii.. 4. Spermatocyte in the growth stage.
Fl<;. 5. Sperm.it ocyte in late prophase.
FK;S. 6 mill 7. Sprrmatocytes in metapliase.
FIG. 8. First spermatocyte anaphase.
Fi<;. '). Second spermatocyte metaphase, lateral view.
FIG. 10. Second spermatocyte anaphase.
FIG. 11. Second oocyte division, metaphase; from Fig. 12.
FIG. 12. Egg in second meiotic division, from a female mated with a male of another species. One sperm is within the egg, at the center, and is swollen; 14 additional sperm are adhering to the outer surface of the egg.
FIG. 13. Haploid metaphase of a cleavage nucleus, from Fig. 15.
FIG. 14. Haploid metaphase of a cleavage nucleus, lateral view, from Fig. 15.
FIG. 15. Anterior portion of an egg in the eighth cleavage division; haploid. Only those cells which occupy that half of the egg nearest the observer are represented in the figure.
SEX RATIO IN MELITTOBIA CHALYBII
PLATE I
261
262 RUDOLF G. SCHMIEDER
separated from the deutoplasm of the egg by a thin zone of clear protoplasm (Fig. 20).
In Figs. 16 to 20 representative portions of a complete ovariole are shown, including those from which the nuclei, drawn on a larger scale in Figs. 21 to 25, were selected.
CLEAVAGE IN UNFERTILIZED EGGS
t "nmated females deposit only a very few eggs and of these few only a small percentage undergoes development. It is therefore more practicable to study haploid development in eggs obtained from females mated with another species of Melittobia, since after such mating ovi- position occurs often at a rate comparable to that occurring in normally mated females. The foreign sperm is generally observed as having entered the egg but actual syngamy seems never to occur, or, at least, never to yield viable zygotes. The small percentage of eggs that even- tually hatches always produces haploid males of the maternal type.
Eggs fixed ten minutes after oviposition are commonly in the second oocyte stage, as shown in Fig. 12. Here is seen the foreign sperm which is within the egg and is considerably swollen, as well as an unusual number of additional sperm adhering to the outside surface of the egg.
PLATE II
FIG. 16. Cephalic end of an ovariolc, with undifterentiated cells, oocytes in synizesis, and post-synaptic spireme; nurse cells and follicle cells.
FIG. 17. Portion of an ovariole with young, growing oocyte and nutritive chamber.
FIG. 18. The same, somewhat older.
FIG. 19. Anterior end of a half grown egg, with the follicle cells.
FIG. 20. Anterior end of a fully grown egg within the ovariole.
FIG. 21. Undifferentiated diploid cell from the ovariole.
FIG. 22. Oocyte nucleus in post-synaptic spireme stage.
FIG. 23. Oocyte nucleus, later stage in early growth period.
FIGS. 24 and 25. Oocyte nuclei during middle growth period; the chromosomes hazy, diffuse, usually more or less ring-shaped.
IK;. 26. Oocyte nucleus preparing for the first meiotic division, from Fig. 27.
IK.. 27. Egg, less than 4 minutes after oviposition; the sperm is slightly enlarged and near the center, the oocyte nucleus near the periphery.
IK;. 28. Egg chromosomes at the end of the first oocyte division, from Fig. 29.
FIG. 29. Egg, less than 15 minutes old, with first meiotic division completed. The sperm is now swollen and at the center of the anterior part of the egg; the egg chromosomes are near the periphery, those of the polocyte nearest the egg surface.
FIG. 30. First cleavage prophase of a fertilized egg, from Fig. 31; two chromo- some groups of five each are distinguishable.
FIG. 31. Egg in prophase of first cleavage; the polocytes are at surface.
FIG. 32. First cleavage metaphase of a fertilized egg, from Fig. 33.
FIG. 33. Egg in metaphase of first cleavage; the polocytes, optically superposed, are at the surface.
FIG. 34. A metaphase from the blastoderm; from caudal end of egg in Fig. 35.
FIG. 35. Cephalic end of egg with young blastoderm; the yolk nuclei are in the center; the blastoderm cells, in this region, are all in anaphase.
SEX RATIO IN MELITTOBIA CHALYBII PLATE II
263
264 RUDOLF G. SCHMIEDER
The 5 chromosomes of the second oocyte and those of the first polocyte are also present and are distinguishable by their relative positions, the oocyte chromosomes being nearer the center of the egg while those of the polocyte are near the surface. Both are shown on a larger scale in Fig. 1 1 .
After completion of the second oocyte division, normal, haploid cleavage may set in in a small percentage of the eggs; but it is not at present known whether such normal cleavage can actually occur in eggs containing a foreign sperm or whether it is confined to eggs which may have escaped penetration by a sperm. In any case, whenever cleavage proceeds normally the nuclei are invariably haploid. The first cleavage mitosis occurs in the center of the anterior half of the egg. During the succeeding cleavages, the mitotic stages are strictly syn- chronous up to the time of formation of the complete superficial blasto- derm, 12 hours later, when the divisions at the hind end of the egg begin to lag slightly behind those at the anterior end. Figure 15 repre- sents the anterior end of an egg in which the cells are all in metaphase of the eighth cleavage division. Polar and lateral views of typical cells from this egg are also shown on a larger scale (Figs. 13 and 14), with the slender, elongated chromosomes characteristic of these divi- sions. In germ-band stages, in late embryos and in young larvae hatch- ing from such eggs only the haploid number of chromosomes has ever been found.
DEVELOPMENT OF FERTILIZED EGGS
In normally fertilized eggs the meiotic stages are like those just described for unfertilized eggs. The sperm is most often found at first caudad to the oocyte nucleus whence it proceeds to the region where the first cleavage mitosis is to occur. Of the ootids, the one which is to take part in syngamy also approaches this region, while the polocytes come to lie at the periphery of the egg. In an egg less than four minutes old (Fig. 27), the sperm head is slightly swollen and the oocyte nucleus (shown on a larger scale in Fig. 26), now shaped like a truncated spindle, is also swollen and is approaching the periph- ery. Figure 29 shows the end of the first meiotic division as well as the enlarged sperm; and Fig. 28, the two chromosome groups more highly magnified. The prophase of the first cleavage mitosis (Figs. 30 and 31) contains 10 chromosomes and it is possible to distinguish two groups of five chromosomes each, one group derived from I he male pronucleus, the other from the female. This stage is followed by a metaphase (Figs. 32 and 33), of which only the lateral view is obtain- able in an egg which has been mounted entire. The succeeding cleav- age divisions and the later stages of development are similar to those
SEX RATIO IN MELITTOBIA CHALYBII 265
described above for unfertilized eggs, excepting that the diploid number of chromosomes is present. Figure 35 represents an egg in which the cleaveige nuclei have just entered the periplasm to form the completed blastoderm. Embedded in the yolk are a number of chromatic masses, the yolk nuclei, while at the surface, each blastoderm cell is in anaphase, excepting those cells at the extreme caudal end of the egg which are still in metaphase. One of these metaphases is shown on a larger scale in Fig. 34.
DISCUSSION
It is apparent that the males in this species are haploid and develop from unfertilized eggs of either mated or virgin females, while the females are always diploid zygotes. Sex determination thus appears to follow the Dzierzon Law except that, in Melittobia, haploidy generally has a lethal effect and only a small proportion of the unfer- tilized eggs is capable of development.
There is nothing, however, in our data that would exclude the validity of the Whiting theory of sex determination. In attempting to apply the latter theory here, we should have to note again that hap- loidy is lethal in most eggs and that there is as yet no evidence for the survival of zygotes produced by homeosyngamy, i.e., for biparental males.
Biparental males may be looked for only among the offspring of normally mated females. Not more than four such males were exam- ined cytologically. The remainder of the males examined were all obtained from virgins or from females mated with males of a different species. Although the direct evidence against the occurrence of biparental males is inadequate, the fact that the number of eggs yield- ing males, compared with the total number of eggs laid, is no larger (ca. 3 per cent) in females normally mated than in females not so mated would seem to suggest that biparental males are not produced. We may therefore conclude that homeosyngamy, if it occurs, is lethal. Without the presence of distinguishable X and Y chromosomes the existence of invariably lethal homeosyngamous combinations is not demonstrable. If such combinations do occur, we should expect to encounter a corresponding number of inviable eggs in the layings of normally mated females. This one condition, at least, is actually met. There is normally a considerable egg mortality observed, which seems to vary in different individuals. In the only instance in which an accurate count was made it amounted to 21 per cent out of the first 224 eggs laid.
The fact that the ratio of males to females (3 : 97), produced by
266 RUDOLF G. SCHMIEDER
normally mated females, is approximately the same as the ratio of males to unhatchable eggs produced by females not so mated, brings to mind an idea expressed by Lenhossc'k and Godlewski. As expressed by Godlewski, this idea is that in honey bees two sorts of eggs are pro- duced, of which only the one sort, the female-producing, is capable of and requires fertilization, while the other produces males partheno- genetically. Although it is now generally recognized that in honey bees every egg is capable both of parthenogenesis and of fertilization, the evidence in Melittobia appears to favor Lenhossek's idea. How- ever, unless we can state more definitely the nature of the differences between the postulated two sorts of eggs, any hypothesis based on this idea remains rather gratuitous and does not contribute to any real understanding of the mechanism responsible for the unusual sex-ratio and the related reproductive phenomena observed in this species. In the material examined there have not been recognized, as yet, any visible differences in the chromosomal behavior during oogenesis or later, which might be considered responsible for the failure of haploid development in most eggs and its success in some.
LITERATURE CITED
DZIERZON, J., 1845. Ueber die Fortpflanzung der Hienen. Eichstatt. Bienen
Zeitung, 1: 113. (Cited from P. \Y. Whiting, 1935). GODLEWSKI, E., 1910-1914. Physiologic der Zeugung. In Winter stein's Ilandb. d.
vergl. Physiol., 3 (2): 457. LENHOSSEK, M. v., 1903. Das Problem der geschlechtsbestimmenden Ursachen.
Jena. (Cited from E. Godlewski, 1910-1914.) ScHMiEUiiK, R. G., 1933. The polymorphic forms of Melittobia chalybii Ashmead
and the determining factors involved in their production (Hymenoptera:
Chalcidoidea, Eulophida-). Biol. Bull., 65: 338. \\ HiTixc;, P. \V., 1935. Sex determination in bees and wasps. Jour. Ilcred., 26: 263
THE RELATION OF FAT CHANGES TO THE GENERAL CHEMICAL EMBRYOLOGY OF THE SEA URCHIN
FREDERICK RONALD HAVES
(From the Zoological Laboratory, Dalhousie University, Halifax, Canada, and the Marine Biological Laboratory, Woods Hole, Mass.)
INTRODUCTION
Several years ago (Hayes, 1934) a study was made of the nitrogen metabolism in the early egg stages of the Florida sea urchin, Ec.hino- metra lucunter. In the previous year Ephrussi published a general ac- count of the metabolism of developing Paracentrottis eggs. The papers agreed in rinding a loss of nitrogen during later development; for the period immediately following fertilization there are no data in the latter paper, and in the former the results suggested an increase in nitrogen which, since such a state of affairs is otherwise unknown in embryology, was treated with caution pending confirmation. Nitro- gen then, and therefore protein, is evidently a source of energy in later development, probably from the eight-hour hatching stage on.
This paper records the results of a study of another possible source of developmental energy, namely fat. Ephrussi (1933) offers the only earlier study, and he found a diminution of fat as development pro- ceeded. Since he tested only three ages, namely unfertilized, 12 hours and 40 hours, his results do not show the precise time at which the loss took place, and cannot be used to supplement and check those to be presented below.
Ephrussi expresses his results as percentages of the wet or dry weight of the egg, a procedure which can be standardized when one is dealing with unfertilized eggs alone, but which presents great difficulties when it is necessary to take into consideration the changes in volume and shape which occur during development. Eggs usually have to be concentrated for analysis by centrifugation, and the number of eggs which will occupy 1 ml. in a centrifuge tube at a given speed is not the same at any two ages. If eggs are sectioned and examined it is possible to derive formulae for the estimation of volume at different stages (Pelluet, 1938). If the volumes are now combined with counts showing how many eggs can be packed into 1 ml., data are at hand for the expression of any egg constituent as a percentage, e.g. fat. This procedure would be very tedious; has not been carried out by any
267
268 FREDERICK ROXALD HAYES
investigator; and if it were would not be profitable because the truth would be obscured by fluctuations in the water content which would make e.g. a stationary fat concentration per egg appear to vary. For these reasons and others which have been discussed at length (Hayes, 1934), the values below are given per million eggs rather than in per cent.
METHODS
The first requirement in work of this kind is a method for rearing eggs in large numbers, free from debris and bacteria, and showing a high percentage of normal development. Attempts were made at first to work with starfish eggs, but the last condition could not be satisfactorily met, and Arbacia was substituted. Eggs were obtained from several females, fertilized and reared in finger bowls according to the well-known methods devised by Just and others. All stages were examined to ascertain the percentage developing normally and it was always over 90.
When it was desired to collect a sample the eggs were concentrated in centrifuge tubes at low speeds. This procedure could not be used for swimming stages, which were collected by filtration through fine bolting silk. Following either treatment, eggs or larvae were trans- ferred to a volumetric flask of sea water, usually of one liter capacity, and the water made up to the mark. Water and eggs were then trans- ferred to a larger vessel and thoroughly mixed. While the mixing was continued 1 ml. samples were withdrawn with a brass Stempel pipette and transferred to a Sedgwick-Rafter plankton counting slide of capacity 1 ml., provided with a cover slip. (I am indebted to Dr. C. J. Fish for the loan of a Stempel pipette and counting slide.) Ten lidds of the slide, selected at random, were then counted with a microscope whose lenses had previously been calibrated with a stage micrometer. The total wras added up and constituted one observation. Usually four such observations were made on each of two samples and the results averaged. Two observers alternated on the counting in order to eliminate the personal factor as far as possible. The probable errors in counting ranged from ± 0.39 per cent to ± 10.2 per cent with an average value of ±4.1 per cent.
After the counts were made the eggs were again concentrated in centrifuge tubes, and transferred from these to a 100 ml. volumetric flask. The fats were extracted by the wet alcohol-ether method described by Fowweather (1926), the flask was made up to the mark, and the fat-containing extract filtered off and stored in glass-stoppered bottles, evaporation being prevented by sealing the tops of the bottles
FAT CHANGES AND KMBRYOLOGY OF SEA URCHIN
269
with several coats of silica (water glass). When the analyses were made some months later, portions of the alcohol-ether extract were dried in a partial vacuum at room temperature and re-extracted with petroleum ether. These were made up to known volume and stored in cork-stoppered bottles, sealed by silica.
The total fat or phospholipid or sterol digitonide, was estimated by a micro-method devised by Backlin in 1930, described by Peters and Van Slyke in 1932, and slightly modified by Van Slyke, Page and Kirk (1933). Carbon dioxide, produced by the combustion of a fat sample in a special test tube attached to a Van Slyke volumetric or manometric apparatus, is carried over and reacts with dilute NaOH in the chamber, being subsequently liberated by the addition of excess
TABLE I
Variations during development in total fat, sterol and phospholipid per million eggs.
|
1 Age in hours and minutes |
2 Total fat, mg. per million |
3 Sterol, mg. per million |
4 Phospholipid, mg. per million |
|
Unfert. |
5.65 |
0.430 |
2.57 |
|
1. |
3.98 |
0.413 |
2.53 |
|
4.40 |
3.50 |
— |
— |
|
6.30 |
3.36 |
0.413 |
3.3 |
|
8.50 |
3.04 |
— |
0.81 |
|
10.50 |
3.60 |
— |
2.04 |
|
15.35 |
3.80 |
— |
2.48 |
|
19.30 |
3.92 |
0.431 |
— |
|
23.15 |
3.84 |
— - |
3.46 |
|
24.50 |
4.29 |
— |
— |
|
25.25 |
3.25 |
— |
1.2 |
|
43.10 |
2.11 |
0.416 |
— |
lactic acid, and measured as a gas. The probable errors resulting from extractions and fat determinations are less than ± 2 per cent. For conversion of the CO2 into fat or into sterol the factors are given in Peters and Van Slyke, page 437. A similar table was constructed for phospholipids, based on data in Bloor (1929).
Sterol was precipitated with digitonin directly from an aliquot of the original alcohol-ether extract in one of the combustion tubes. The precipitate was washed with ether and with water, supernatent fluid being removed by means of suction through an alundum filter stick as described by Kirk, Page and Van Slyke (1934). After drying, the sterol digitonide was oxidized to CO2 as described above. The method was checked and found to give theoretical values with cholesterol.
270
FREDERICK RONALD HAVES
TOTAL FAT
By total fat is meant all the material soluble in petroleum ether. Its variations are given in Table I, column 2 and in Fig. 1. It will be noted that from an initial value of 5.65 mg. per million in tin- unfertilized egg, there is a steady drop until by 9 hours there are only 3.04 mg. From this time until 20-25 hours an increase is noted, followed by a drop to the final stage observed. The most interesting points are that the trough of the curve coincides with the time of hatching, and that there is evidence of fat synthesis from approxi- mately 9 to 23 hours. The shedding of the shell and the beginning of fat synthesis may well be associated with one another. No previous
10 15 20 25 30 Hours After Fertilization
35
40
Fit;. 1. Fluctuations in total fat during development. Note that the time of hatch- ing, 8 hours, corresponds to the initial trough of the curve.
reference to the manufacture of fat by invertebrates has been found, although the phenomenon is not unknown in teleost embryos (Hayes and Ross, 1936). Since the embryo has at this time no organs of digestion, the source of the new fat is presumably some other material in the egg. There is no proof as to what this is, but a hint is obtained from the nitrogen metabolism of the Florida sea urchin, Echinometra (Hayes, 1934), in which, at about the time of hatching a decline in the non-amino nitrogen began, which was practically arrested when the observations closed at 24 hours. (The amino nitrogen, on the other hand, did not begin to decline until the sixteenth hour.)
FAT CHANGES AND EMBRYOLOGY OF SEA URCHIN 271
Ephrussi's three values for total fat were as follows, given in each case as percentage of the wet weight.
Unfertilized 4.81
12 hours 4.43
40 hours 3.69
A glance at Fig. 1 will show that the peak at 20 hours will be missed by the particular stages selected by Ephrussi, so that his gentle decline is in agreement with Fig. 1 .
Three further pieces of evidence may be mentioned, after which it will be possible to present a provisional picture of the general chemistry of early development. Warburg (1915) found in Strongy- locentrotus a sudden burst of respiratory activity at fertilization, followed by a steady increase as development proceeded. The respiratory quotient over the first three hours was 0.9.
Secondly Runnstrom (1933) observed in Paracentrotus a sudden production of acid at fertilization. It was not carbonic acid — in fact the pH of the egg was altered to such an extent that considerable quantities of CO2 were driven off. The consequence was that Runn- strom, who was making an investigation of the changes in respiration during the first hour, found that the excess CO2 caused the apparent R.Q. for the first 10 minutes to soar to 2.37. Within half an hour the mysterious acid was being used up, and respiratory CO2 conse- quently held in the egg as bicarbonate, with the result that the ap- parent R.Q. for 30-45 minutes was only 0.45. Although Runnstrom did not discover what the acid in question was, he showed that it was not phosphoric acid, and he cited the work of Perlzweig and Barron on Arbacia to prove that it was not lactic acid. The inference is that carbohydrate metabolism is not responsible. It may be noted here for future reference that Page (1927) found that the oil of Arbacia eggs had a saponification value of 606; of Asterias eggs 319. The oils also yielded large quantities of volatile fatty acid when distilled with steam. Fatty acids from such oils would have a low molecular weight, high solubility in water, and adequate strength to drive out CO2.
Finally, evidence has been accumulating for some years which suggests that two kinds of respiration go on simultaneously in the egg (see Loeb and Wasteneys, 1911). It is not improbable that two substrates are burnt and that two enzymes facilitate the oxidations. That one of these belongs to the general group of oxidases, peroxidases and catalase is shown by the fact that the respiration both of fertilized and unfertilized eggs is stimulated by dimethylparaphenylene diamine (Orstrom, 1932). This reagent is absorbed by the eggs and becomes
272 FREDERICK RONALD HAYES
in effect an artificial substrate, the measure of whose oxygen require- ment is an index of the available excess enzyme in the egg. The enzyme system then, is present before fertilization, but of course the amount of oxidation which it can bring about is strictly limited by the quantity of available substrate. A second characteristic of the oxidase type of enzyme is its susceptibility to cyanide poisoning. Now cyanide has very little effect on the respiration of the unfertilized egg (Runnstro'm, 1930), but it produces a marked inhibition of oxygen consumption after fertilization. The conclusion is that there is an abundance of enzyme, but practically no substrate before fertilization, and a small but steady supply after fertilization, which is used up as produced. As to the second part of respiration (not inhibited by cyanide) it was found that it could be stopped by narcotics, e.g. lithium (Lindahl, 1934), and stimulated by pyocyanine (Runnstro'm, 1935a, both fertilized and unfertilized eggs), and by methylene blue (tried on unfertilized eggs only by Barren, 1929). Thus it passed the tests for a dehydrogenase system. Pyocyanine acts as an artificial enzyme, and the stimulus caused by its presence meant that there must have been some extra substrate there to be burnt up; in other words (in contrast with the first part) the substrate was plentiful and the small quantity of enzyme was the limiting factor.
Taking all these facts into consideration, assuming that the various sea urchins resemble one another in the essential features of their embryonic metabolism, and remembering that the times of hatching approximately coincide, the following general hypothesis of echinoid chemical embryology may be advanced.
(1) At the time of fertilization a considerable quantity of fat is split into fatty acid and glycerol, presumably by means of a lipase, the activation of which is the first of the chain of events initiating embryonic metabolism. It would be tempting to believe that the lipase is brought in by the spermatozoon, but no evidence has been found bearing on this point. For every molecule of glycerol produced there are three molecules of fatty acid; and for every calorie of energy available from glycerol there are 5 or 10 calories available from a lower fatty acid. The glycerol is scarce, the acid plentiful.
(2) The glycerol is rapidly oxidized by a cyanide-sensitive rn/yme. This part of respiration does not increase as development proceeds. The enzyme was there before fertilization, but was inactive because of the absence of suitable substrate. The destruction of glycerol accounts in part for the increase of activity found at fertilization by Warburg, (ilycerol continues to be produced in small quantities up to at least 8 hours, bin is oxidized as formed. That cell division is
FAT CHANGES AND EMBRYOLOGY OF SEA URCHIN
the part of development presided over by cyanide-sensitive respiration was shown many years ago by Loeb and Wasteneys (1911), who found that the concentration of cyanide exactly necessary to block development reduced the oxygen consumption to one-quarter its normal value, an amount which coincided with the normal requirement of unfertilized eggs. Runnstrom (1935a) has recently confirmed this experiment, and extended it by immersing eggs in a mixture of cyanide and pyocyanine at the same time, thereby inhibiting one part of respiration and stimulating the other. The net result was an increase in oxygen consumption, but a blocking of cell division in early prophase.
(3) Runnstrom's unknown acid is fatty acid. It is oxidized by the second, or cyanide-insensitive dehydrogenase system. The ac- tivities resulting from this energy source are presumably growth and basal metabolism. The enzyme does not oxidize all its substrate at once — if it did the egg would burn out in a very short time. Never- theless, this part of respiration is shown by experiment to increase in intensity as development proceeds. An explanation of the paradox of how a limited quantity of enzyme can be made to do more work later in development than at the beginning is obvious from the work of Spek (1934), who showed that in Asterias and Arbacia eggs acid is not distributed evenly throughout the whole egg, but is strictly local- ised. Before fertilization the surface is acid, the interior alkaline; in early developmental stages there is a thin acid layer surrounding the dividing cells, and from the 64-cell stage to the gastrula there is a gradient of intensity of acid reaction between the animal and vegetal poles. The chemical geography of the egg limits the extent to which enzyme and substrate are able to come together. This conception is strengthened by Runnstrom's (19356) statement that "All our experi- mental evidence indicates that the susceptibility to the action of lithium is highest at the animal pole and decreases gradually."
(4) Fat ceases to be a source of energy at approximately the time of hatching, 8 hours, possibly because it is required as building material.
(5) From hatching time up to 16 hours energy is provided by non-amino nitrogenous materials (non-protein?), from which fat is synthesized and possibly the costs of cell division, motion and basal metabolism are in part or in whole met.
(6) To the utilization of non-amino nitrogenous materials there is added at 16 hours, a gradual destruction of compounds containing amino nitrogen as well.
(7) At 24 hours fat begins once again to decline while at the same time the destruction of nitrogenous material appears to be concluding.
274 FREDERICK RONALD HAYES
(8) Ephrussi (1933) reports the loss of almost all the carbohydrate in the egg between his 12- and 40-hour tests, and very little loss in the first 12 hours. This suggests that carbohydrate is the fuel source drawn upon in later stages, possibly by about 24 hours. The embryo is perhaps using up its last reserves of food and consequently entering upon a period of starvation, pending the development of its digestive system to a point where food can be taken from the sea.
(9) Hence the apparent sources of energy in succession are: fat, non-animo nitrogen, amino nitrogen, fat, carbohydrate. More de- tailed information may move carbohydrate to the second or third last place.
The assignment made above of cyanide-sensitive enzyme to glycerol (or a product of it), and of cyanide-insensitive enzyme to fatty acid, is based not only on the relative abundance of the substrates, but also on the work of Emerson (1927), which indicates that cyanide-sensitive respiration results from the oxidation of carbohydrate.
Lindahl and Ohman (1936) have recently placed on some of the results cited above an interpretation differing from that given here. They say, "We see a useful conception of these relationships in the assumption that two different substrates are burnt, of which one has a small, the other a high degree of reactivity to the oxidizing 'carrier.' Shortly after fertilization only the first substrate is present in large quantities. The unchanging and small concentration of the 'carrier* determines the constant and small intensity of oxidation of this substrate. This reaction corresponds to the non-growing part of respiration, which can be increased to a marked degree by dimethyl- paraphenylene diamine and pyocyanine. The 'carrier' is here the limiting factor. The other substrate is very rapidly oxidized and hence does not accumulate. The rapidity of formation of this sub- strate determines the oxygen consumption, and is therefore the limiting factor. This is the 'growing part' of respiration, which can be inhibited by lithium." It will be seen that Lindahl and Ohman difter from the views expressed in this paper in that they: («) postulate two substrates but only one enzyme system; (b) express no view as to the nature of the substrates; (r) consider that dimethylpara- phenylene diamine and pyocyanine act on the same part of respiration; . (d) state that the cyanide-sensitive part of respiration is limited by the low degree of reactivity of the substrate (and not by its scarcity); (e) state that the cyanide-insensitive part of respiration is limited by the scarcity of the substrate1 (and not by geographical segregation)-
FAT CHANGES AND EMBRYOLOGY OF SEA URCHIN
STEROL
The values for sterol will be found in column 3 of Table I. It amounts to some 0.41-0.43 mg. per million eggs, and no fluctuation was found in its concentration during the period investigated. The only comment necessary is that this particular one-tenth of the fat is not a source of embryonic energy.
PHOSPHOLIPID
The results of phospholipid determinations are given in column 4. They vary from less than 1 mg. to nearly 3.5 mg. per million eggs. The values unfortunately fluctuate so much that attempts to discover a developmental trend by plotting them on a graph are unconvincing. No utilization of phospholipid can be said to be demonstrated, although the data are not sufficiently extensive to exclude the possibility. The average of all the phospholipid results is 2.17 ± 0.24 mg. per million eggs or 38 ± 4 per cent of the total fat in the unfertilized egg.
PREVIOUS ANALYSES
McClendon (1909) found that the total fat (ether extract) of unfertilized Arbacia eggs was 2.254 per cent. Harvey (1932) gives the data necessary to convert this to mg. per million as follows:
Mg. per million eggs =
106 X vol. of 1 egg in cu. mm.
X egg density X 2.254
100
= 2.12 X 1.09 X 2.254 = 5.21 mg.
McClendon also estimated the phosphorus content of his ether extract and found it to be 0.06914 per cent of the whole egg. Assuming that all the phosphorus is in the phospholipids and that it comprises 4 per cent of these (as it does in lecithin) then this figure too can be made comparable.
2.12 X 1.09 X 0.06914 X 100 ~4~
= 3.84 mg. phospholipids per million eggs.
If it is valid to apply these calculations to McClendon's results then it follows that they agree in a general way with those given above.
Page (1927) has also made analyses of the fatty constituents of the unfertilized Arbacia egg. He states that 8.3 grams of oil were obtained from 183 million eggs by means of alcohol-ether extraction. This is in the terminology of this paper, 45.4 mg. per million eggs, or
276 FREDERICK RONALD HAYES
ten times as great as McClendon's figures and those given here. A calculation of the mass of an egg from the volume and density values in Harvey shows that, according to Page, 20 per cent of the egg is fat. Now since the total solids (McClendon) only comprise 20 per cent of the egg and these are obviously not all fat, Page's results are difficult to interpret, particularly in view of his statement that Arbacia contains much less oil then Asterias. Page also precipi- tated the phospholipid from his extract with acetone, and his results work out at 8.4 mg. per million as against McClendon's 3.84 mg. and the 2.17 mg. above. Finally Page distilled the volatile fatty acids from his extract with steam and titrated them with NaOH, finding 9.4 mg. per million eggs. The discrepancy between Page's results and the others may lie in pan in his method of counting eggs, no mention of which is made in his paper. This difficulty would not invalidate his estimates of the saponification value and his volatile acid determinations on which a part of the general argument in this paper depends.
SUMMARY
Periodical estimations were made of the total fat, sterol and phospholipid during the first 40 hours of development of Arbacia. The total fat decreases up to the time of hatching (8 hours), then increases for some 10 hours, and later decreases again. The sterol concentration remains unchanged throughout the period studied. Owing to fluctuations in the phospholipid readings, a definite con- clusion could not be drawn as to whether this material is utilized as a source of embryonic energy. The total fat in a million unfertilized eggs is 5.65 mg. of which 7.5 per cent is sterol and 38 per cent phospho- lipid. A general hypothesis of the chemistry of early sea urchin development is presented.
I ( is a pleasure to acknowledge my indebtedness to Professor E. (j. Young, whose advice had been frequently sought and most generously given.
LITERATURE
UACKI.IN, E., 1930. Biochem. Zeilsc.hr., 217: 4cS2.
HARRON, E. S. (",., 1929. Jour. Biol. Client., 81: 445.
BLOMR, W. R., 1929. Jour. Biol. Chem., 82: 273.
EMERSON, R., 1927. Jour. Gen. Physiol., 10: 469.
EPHRUSSI, I}., 1933. Arch, de Biol., 44: 1.
FOWWEATHER, F. S., 1926. Brit. Jour. Exper. Path., 7: 7.
HARVEY, E. N., 1932. Biol. Bull., 62: HI.
HAYES, F. R., 1934. Carnegie Inst. Wash. Pithl. No. 435 (Papers from Torlugas
Lal>., 28: 1X3).
HAYKS, F. R. AND D. M. Ross, 1936. Proc. Roy. Soc., London, Series B, 121: 358. KIKK, E., I. II. PAGE AND D. D. VAN SLYKE, 1934. Jour. Biol. Clu-in., 106: 203.
FAT CHANGES AND EMBRYOLOGY OF SEA URCHIN
LINDAHL, P. E., 1934. Naturwiss., 22: 105.
LINDAHL, P. E. AND L. O. OHM AN, 1936. Naturwiss., 24: 157.
LOEB, J. AND H. WASTENEYS, 1911. Biocftem. Zeitschr., 36: 345.
McCLENDON, J. F., 1909. Am. Jour. Physiol., 23: 460.
ORSTROM, A., 1932. Protoplasma, 15: 566.
PAGE, I. H., 1927. Biol. Bull., 52: 164.
PELLUET, D., 1938. Quart. Jour. Micros. Sci. In press.
PETERS, J. P. AND I). I). VAN SLYKE, 1932. Quantitative Clinical Chemistry.
Williams and Wilkins Co., Baltimore. RUNNSTROM, J., 1930. Protoplasma, 10: 106. RUNNSTROM, J., 1933. Biocliem. Zeitschr., 258: 257. RUNNSTROM, J., 1935a. Biol. Bull., 68: 327. RUNNSTROM, J., 19356. Biol. Bull., 68: 378. SPEK, J., 1934. Protoplasma, 21: 561.
VAN SLYKE, D. D., I. H. PAGE AND E. KIRK, 1933. Jour. Biol. Chcni., 102: 635. WARBURG, O., 1915. Arch. f. d. ges. Physiol., 160: 324. For a translation of this
paper see Warburg, O., 1930. The Metabolism of Tumors. Constable
and Co., London.
ON THE SPECIFICITY AND RELATED PROPERTIES OF THE CRUSTACEAN CHROMATOPHOROTROPIC
HORMONE
A. A. ABRAMOWITZ AND R. K. AHRAMOWITZ (From the Marine Biological Laboratory, Woods Hole, Massachusetts)
INTRODUCTION
\Yhen tlie eye-stalks of several brachyuran crustaceans are removed, the animals become pale within two hours (1), (2), (3), (4). Injection of a sea-water extract of the extirpated eye-stalks into the blinded (eye-stalk-amputated) specimens results in the appearance of the dark coloration (3), (5). Although several kinds of chromatophores are involved in these color changes, the movements of the melanophores are chiefly responsible for the resulting external coloration, and in I In- following experiments were the only chromatophores studied in the fiddler crab, Uca pugilator.
RESULTS Method of Determining the Relative Activity of Eye-stalk Extracts
A new method of assaying the eye-stalk hormone was devised so that determinations could be made within an hour. The techniques of preparing the hormone, injecting the animals, and preparing the animals for the test were essentially the same as described previously (5). The plan of the method was as follows: Each of 8 groups of 15 blinded animals per group was injected with one of 8 different, known concentrations of hormone (1.0 E.S.1— 0.005 E.S. per cc. of solution). The percentage of animals showing the slightest perceptible response (melanophore stellation) was determined at 5-minute intervals follow- ing the time of injection for a period of one hour. If the percentage of animals showing the slightest perceptible response at various concen- trations is plotted against time, a series of steep, sigmoid curves is obtained whose inflection points intersect the ordinate at about 50 per cent. The times at which 50 per cent of the animals show tin- response at different concentrations range from S minutes for the strongest to 24 minutes for the weakest. A standard curve (Fig. 1) was then constructed, using a range of concentrations from 0.04
'The letters E.S. are the abbreviation for eye-stalk. The letters E.S. II. will be used as the abbreviation of eye-stalk hormone.
278
CHROMATOPHOROTROPIC HORMONE OF UCA
279
E.S./CC.-0.002 E.S./cc. Figure 1 can, therefore, be utilized for deter- mining the relative strength of very dilute solutions of the hormone. For stronger concentrations (1.0 E.S./cc.-0.03 E.S./cc.) Fig. 2 was employed, indicating the relationship between concentration and percentage of animals showing the full response (complete melanophore expansion) at 1 hour following the time of injection. For still stronger solutions, the method previously described (5) was employed. By
FIG. 1. Curve showing relationship between percentage of animals showing the slightest perceptible response (melanophore stellation) and concentration. Ordinate — percentage of animals responding; abscissa — concentration expressed as eye-stalks per cc. of solution.
using Figs. 1 and 2 as characteristic curves, the relative strength of an unknown concentration can be determined by taking a reading, at 30 minutes after injection, of the percentage of animals showing the slightest perceptible response, and another at 1 hour giving the percentage showing the full response.
A study of these and other curves not included in the text reveals
A. A. ABRAMOWITZ AND R. K. ABRAMOWITZ
several points of interest concerning the nature of melanophore re- sponse in Uca. From the results shown by Fig. 1, where very small amounts of hormone were injected, it is evident that a certain threshold must be reached before the melanophores react and, moreover, that this threshold varies from animal to animal. With stronger doses, the
100
90
80
70
60
50
40
30
20
10
.1
.3
.4
.8
.9
FIG. 2. Curve showing the relationship between percentage of animals showing the full response (complete melanophore expansion) and concentration of hormone. Orclinate — percentage of animals showing full response; abscissa — concentration expressed as eye-stalks per cc. of solution.
degree and the duration of melanophore expansion are exponentially proportional to the dosage injected. Thus, melanophores, like certain types of smooth muscle, show graded responses whose magnitude varies as the strength of the stimulus. For a complete expression among time, concentration of hormone, and the various phases o! melano-
CHROMATOPHOROTROPIC HORMONE OF UCA 81
phore response, the construction of several nomograms is necessary. The curves already presented, however, suffice for the present studies.
Sensitivity of the Test; Estimation of the Amount of Hormone in the
Eye-stalk
The strength of eye-stalk extracts, as read off from the curves, is not constant from animal to animal. Those from larger animals (5.0 grams) usually assay at higher values, although there is always some discrepancy in a group of the same size and sex. The sensitivity of the test may be illustrated by the following experiments: a sea- water extract of 1 eye-stalk (animal weight = 5.0 grams) was diluted until the total volume was 400 cc. Sixty-six per cent of the animals injected with 1/20 cc. of this solution responded. When further diluted to 600 cc., 50 per cent responded; to 800 cc., 33 per cent responded. With still greater dilutions, the percentage quickly fell to zero. We take the point at which 50 per cent of the animals respond as a reliable indication that an extract is active, and in the following discussion, we assume this point (50 per cent response) to indicate a minimal unit of activity, which is 1/600 of a Uca unit (5).
The dry or nearly dry weight of an eye-stalk is 2.0 mg. (average of 600 eye-stalks). From microscopic sections of the eye-stalks of Uca pugilator, we estimate the sinus gland of Hanstrom (6) to be, roughly, 1/100 that of the eye-stalk. Assuming the proportion of the active principle to be 1/100 that of the gland, which is probably a conservative surmise, the total amount of hormone in one eye-stalk is therefore about 0.2 7. The sensitivity of this biological test is such that 0.2 -T- 600 X 20 (since each animal is injected with 0.05 cc.) or 0.000016 7 of hormone can be detected. These figures, of course, are very rough but, we think, indicate the probable order of magnitude.
Specificity of A ction
Thus far, extracts of only two other glands, the pituitary (7) and the subneural gland complex (8), have been found to produce melano- phore expansion when injected into blinded Uca. We found acci- dentally that the injection of distilled water produces strong melano- phore responses. In 55 crabs, blinded for from 2-8 days previously, the injection of distilled water produced in 15-20 minutes complete melanophore expansion in 90 per cent of the animals. After 1 hour, 50 per cent became pale, and after 2 hours, 90 per cent-100 per cent
A. A. ABRAMOWITZ AXD R. K. ABRAMOWITZ
became pale. Quantitatively, this effect is quite different from that produced by the eye-stalk hormone in any concentration. Injection of distilled water into 25 normal dark animals was without effect.
Since this reaction is of interest in relation to the general problem of hormone specificity, we examined several possibilities which might explain the effectiveness of distilled water. It is quite certain that injection of sea water does not produce any responses. Only 1 of 46 blinded animals injected with sea water responded weakly but posi- tively, while no significant changes were detected in 10 normal dark animals also injected with sea water. Sea water, therefore, must be employed as the solvent when testing preparations of E.S.H. Since distilled water is acidic (pH 5.7) as compared with sea water (pH 7.7), we injected tap water (pH 7.0) into 10 blinded and 10 normal Uca, with results identical to those already described for distilled water. Identical results were also obtained with distilled water made alkaline (pH 8.0) by the addition of 0.01 N NaOH. The acidity of distilled \vater is therefore not responsible for its chromatophorotropic activity.2
Another difference between the action of the hormone and that of distilled water was obtained by injecting various dilutions of sea water with distilled water. The following proportions of sea and distilled water were tried :
Sea Water Distilled Water Chromatophorotropic Activity
1 part plus 1 part Inactive
1 part pin-- 2 part* 60 per cent of animals respond
1 part plus 3 parts 80 per cent of animals respond
1 part plus 6 parts 80-100 per cent of animals respond
1 part plus 10 parts 80-100 per cent of animals respond
1 part plus 16 parts 80-100 per cent of animals respond
Since the injection of 0.05 cc. distilled water must dilute the blood of a crab appreciably, osmotic effects were studied. Assuming tin- freezing point depression of Uca blood to be close to that of lobster blood, we injected sucrose (1.3 M) isosmotic with the latter into 20 blinded crabs. No melanophore expansion resulted. Sucrose (0.95 M) isosmotic with sea water, however, produced a slight reaction in 20 per cent of the animals. It seemed, therefore, that the effect of distilled water may be due to the resulting hypotonic condition ol the blood. The following salts isosmotic with sea water were also tested:
2 The pH of distilled, tap, and sea water at Woods Hole was determined by means of the jHass electrode.
CHROMATOPHOKOTKOPIC HORMONE OF UCA
283
Blinded Specimens (20 animals pci tesl <
|
Salt |
|
|
NaCl |
0.52 M |
|
MgCls |
0.51 M |
|
CaCl2" |
0.34 M |
|
KCI |
0.53 M |
|
Na,HS04 |
0.4 M |
LiCl 0.52 M
200 per cent sea water
General Effects
Muscular twitchings and death Immediate prostration, death
after 1 hour Immediate twitchings of legs
Chromatophorotropic Activit y (per cent of animals responding)
30 20 20 0 10
20 0
While the effect of drugs on vertebrate melanophores has been determined, no similar investigation has been made on crustacean melanophores. Sixteen different, chemically-pure drugs were in- jected into both normal and blinded crabs. One hundred 7 of the drugs, dissolved in sea water, was injected into each of 10-20 animals of both groups. Several drugs produced violent, convulsive move- ments, prostration or twitchings of the legs while others were entirely without muscular effects. Table I summarizes the action of 100 7 of the drugs:
TABLE I
Effects of sixteen drugs on normal and blinded crabs
|
Drug |
Blinded (Pale Animals) |
Normal (Dark Animals) |
||
|
General Effect |
Chromato- phorotropic Activity (Melano- phore expansion) |
General Effect |
Chromato- phorotropic Activity (Melano- phore contraction) |
|
|
Atropine sulphate |
— |
— |
— |
— |
|
Morphine sulphate . |
— |
— |
— |
— |
|
Acetyl choline. . |
— |
20% (slight) |
Shedding of legs |
— |
|
Histamine ... |
Paralysis of legs |
— |
Slight paralysis |
— |
|
Eserine sulphate |
Prostration and death |
— |
Not performed |
|
|
Pilocarpine HCI. |
— |
— |
— |
|
|
Cocaine HCI |
Rigidity |
30% (slight) |
Rigidity |
— |
|
Brucine sulphate |
Paralysis |
— |
Sluggishness |
— |
|
Veratrine sulphate |
Instantaneous death |
— |
Instantaneous death |
50% slightly pale |
|
Curare . . |
Paralysis |
50% (slight) |
Paralysis |
— |
|
Strychnine SO-i. |
Temporary paralysis |
20% (slight) |
Paralysis |
— |
|
Guanidine |
Prostration |
20% (slight) |
— |
— |
|
Chlorbutanol |
— |
20% (slight) |
— |
— |
|
Caffein |
— |
20% (slight) |
— |
— |
|
Nicotine |
Prostration |
40% (slight) |
Prostration |
— |
|
Hyoscine HBr. . . |
— |
80% positive |
— |
— |
284 A. A. ABRAMOWITZ AND R. K. ABRAMOWITZ
None of these drugs \vith the exception of hyoscine hydrobromide produced definite, positive results in blinded specimens, and all were without effect on the melanophores in normal dark specimens. These drugs (with the one exception) have therefore no direct or indirect action on contracted or expanded melanophores. We did not repeat these injections in normal animals during the night (nocturnal or pale phase) although possibly some of these drugs may effect a release of the hormone from the sinus gland, which appears to be under nervous control (5).
Adrenalin, practically universally, produces melanophore con- traction in vertebrates in extremely small doses. Adrenalin, in various dilutions from 1 : 1,000 to 1 : 10,000 (dilutions greater than 1 : 1,000 being made with sea water) was without significant melanophore responses when injected into either normal or blinded specimens. Strong doses usually produced death in normal animals, while 20 per cent of the blinded animals showed slight melanophore stellation. The activity of adrenalin was confirmed by injecting 0.2 cc. of each dilution subcutaneously in normal dark catfishes. This experiment illustrates effectively the point made elsewhere (4) that melanophores of various animals do not always react in the same way to the same substance, for example, chemically pure adrenalin hydrochloride. It is only logical to think of a response to a hormone in terms of the interaction between the responding tissue and the hormone in question, not solely in terms of the hormone itself. If the same type of respond- ing tissue in two different animals is physiologically and anatomically different, as are the melanophores of crustaceans and vertebrates, it is not surprising that they react in different ways to the same substance. The ineffectiveness of adrenalin in contracting crab melanophores is therefore not a puzzling phenomenon. In fact, it is almost surprising that in one case (7) (4), intermedin produces the same response in crab and vertebrate melanophores.
Do Organs Other than the Eye-stalk Produce the Hormone?
This question has been discussed in detail (5) but as yet no con- clusive answer to it has been given. It has been reported (9) that ventral nerve cord extracts of Palxmonetes are slightly active on the chromatophores of blinded Palaemonetes, and therefore that the ventral nerve cord produces the active principle or principles. A criticism of this conclusion is the possibility that the extracts, prepared from normal animals, may have contained traces of K.S.H. present in the blood bathing the nerve cord.
\Ve have extracted several organs of normal pale, normal dark,
CHROMATOPHOROTROPIC HORMONE OF UCA
and blinded crabs (blinded previously for various periods of time), and tested the extracts on crabs blinded previously for from 2-8 days. For purposes of comparing the potency of any organ found to be active with that of the eye-stalk, the minimal unit of activity as already described was employed. The wet weight of an eye-stalk is about 6.0 mg., and consequently 600 cc. -4- 6 mg. or a 0.001 per cent solution of the eye-stalk represents a minimal unit. Stomach, liver, muscle and heart tissue of both normal dark and blinded specimens were extracted by the usual method and prepared in a 0.1 per cent solution; in other words, 100 times more concentrated than that representing a minimal unit of eye-stalk extract. All extracts were found to be completely inactive. Finally, we resorted to extraction of entire, normal dark crabs, immediately after extirpating the eye-stalks. The extract was prepared in a 2 per cent solution, or 2,000 times more concentrated than the minimal unit of eye-stalk extract. When tested, it was found to be 5 times stronger than the minimal unit. Repetition of this experiment with normal pale animals and animals blinded for from 2 days to 1 month gave identical results. We must conclude that either the hormone is present or being produced by some tissue in the body even in the absence of the eye-stalks, or that the positive result is an artifact brought about by the injection of a hypertonic solution. If the former is true, it can be of little or no significance in the normal chromatic physiology of these animals, as shown sub- sequently.
Behavior of Isolated Leg Melanophores
The behavior of isolated scale melanophores of fish to various ions and organic substances has been studied (10) for a long time. One of the most remarkable results is the behavior of Fundulus scale melano- phores to sodium and potassium ions, expanding to the former and contracting to the latter. We have studied the melanophores of crabs to various substances by cutting off the legs with a fine scissors at a point near the articulation of the femur with the body, and im- mersing these isolated legs into various solutions. The results are summarized in Table II, readings being taken every fifteen minutes after immersion in the solutions and continued for 5 hours.
The results are not particularly illuminating. It is significant, however, that the effect of distilled and tap water is similar to that produced when they are injected into the animal. Sea water also acts in the same fashion as when injected into pale animals. The fact that dark legs paled in sea water after 1^ hours may be due to the diffusion, destruction, or inactivation of E.S.H. already present in
286
A. A. ABRAMOVVITZ AND R. K. ABRAMOWITZ
|
MgCIj |
.37 M |
|
Sucrose |
1.30 M |
|
LiCl |
.52 M |
|
Na2HPO4 |
.40 M |
|
NaCl |
.52 M |
|
CaCU |
.34 M |
|
KC1 |
.53 M |
|
Sea water |
|
|
Tap water |
|
|
Distilled |
the legs. This resembles the last phase of a typical response of blinded crabs to the injection of E.S.H., for when the melanophores begin to contract it requires about \}/% hours to attain complete pallor (5). In view of this, and the slowness of melanophore reaction in crabs, we cannot speak of "ionic effects" on melanophores provided the responses are produced well within an hour. Thus, lithium, sodium, potassium, and calcium chlorides all produce a rapid contraction of expanded melanophores, and consequently, there is no difference, as
TABLE II
Legs of blinded crabs (melanophores contracted) Solutions Results of Immersion
No change for 4 hours, then expansion of femur melano- phores
Same as above, although erythrophores expand within 30 minutes
No change
Expansion within 15 minutes
Slight stellation for first hour, then contraction
No change for 4 hours, slight expansion on fifth hour
Slight stellation for 1 ' •> hours, then contraction
No change during 5 hours
Expansion within hour, lasting for 5 hours on femur
Same as above
Legs of normal dark crabs (melanophores expanded)
Results of Immersion
Slight contraction in 15 minutes to 4 hours, then expansion
on femur
Slight contraction in I 2 hour, expansion after 2 hours Contraction in 15 minutes to 5 hours Remain expanded for 5 hours Contraction in 15 minutes to 5 hours
Contraction 15 minutes to 4 hours, then expansion on femur Contraction 15 minutes to 5 hours Contraction after 1 ' 2 hours Slight contraction after 1 } •> hours, but expansion after 4
hours Distilled Same as above
found in fish melanophores, between the effects of NaCl and K('l. Melanophores in the legs of blinded animals remain more or less con- tracted in these solutions for 4 hours, so that the effect of these various cations is the same regardless of the original state of the melanophores. However, since the chloride ion is the common anion to all four salts, the result may well be due to it. \Ye have not pursued this aspect further for it is outside the scope of this paper. Isolated leg melano- phores of crabs are not valuable material for such an investigation. Their responses are slow, and usually irregular. Most frequently the
|
Solutions |
|
|
MgCU |
.37 M |
|
Sucrose |
1.30 M |
|
LiCl |
.52 M |
|
Na2HPO4 |
.40 M |
|
NaCl |
.52 M |
|
CaCh |
.34 M |
|
KC1 |
.53 M |
|
Sea water |
|
|
Tap water |
CIIROMATOPIIOROTKf )!'[(/ HORMONE OF UCA
reaction starts at the cut surface of the femur and progresses inwardly towards the tibia. Occasionally, the melanophores of the entire femur respond simultaneously, and rarely do the pigment masses of the tibia react prior to those of the femur.
Is the Pale State Constant in Blinded Animals?
It was stated previously that if tissues other than the eye-stalk are capable of producing E.S.H. they must play an insignificant part, if any, in the normal chromatic physiology of the animals. The basis for this belief is that animals once blinded remain in a pale state as long as they remain alive. In all of the experiments, 15-20 blinded animals were maintained in large crystallizing dishes containing about }/2 inch of sea water which was changed daily. Under these conditions, blinded animals have been maintained for 2 months, during which they remained entirely pale.
When blinded animals die, they turn quite dark. Even previous to death, the animals turn gradually dark, and remain dark for some time after death. After observing several hundred blinded animals from day to day, we can state that pallor after enucleation of the eye- stalks is largely constant so long as the crabs remain in a healthy state. Thus far, we have not observed regeneration of the eye-stalk, although several puzzling observations, exceptions to the above statement, have suggested this as a possible explanation. The following protocol illustrates the condition of the melanophores in blinded animals:
Aug. 31 — 158 animals blinded
Sept. 1—100% pale
Sept. 2—98% pale; 2% dark (dead)
Sept. 4—1% slightly dark; 99% pale
Sept. 5 — during day 100% pale; during night 100% pale
Sept. 6-13 — animals divided into 4 groups:
white background during day -100% pale white background during night — 100% pale black background during day -100% pale black background during night — 100% pale Sept. 13 — 142 animals remaining: 3% slightly darkl 5% fully dark J ~ 92% pale = B
Sept. 14— A = 45% dead; 55% dark; B == 93% pale Sept. 15— A == 33% dark; 66% pale (day)
100% pale (night)
Sept. 16— A == 84% dark; 16% pale (day) 84% pale; 16% dark (night)
A. A. ABRA.MOWITZ AX I) R. K. ABRAMOWITZ
The protocol illustrates that over 90 per cent of the animals blinded for 2 weeks remain continuously pale. From 2 per cent-4 per cent, at the end of two weeks, show a slight diurnal rhythm. Further proof of this behavior was obtained by placing 40 animals, blinded two weeks previously, in a glass dish near a window and 25 similar animals in a glass dish in total darkness.
Exposed to Normal Day and Night Conditions
Sept. 14 I'.M. . 95% P; 5% slightly 1>
Sept. 15 A.M. ..95%P; 5% slightly 1)
Sept. 15 P.M. ... 95% P; 5% slightly D
Sept. 16 A.M. 70% P; 30% slightly D
Sept. 16P.M. 97% P; 3% slightly D
Sept. 17 A.M.. . 67% P; 33% slightly 1)
P = pale. I) = dark.
Darkness
100% P 100% P 100% P 100% P
ioor; p 100% P
A certain percentage of blinded animals is therefore capable of undergoing periodic changes in color. Usually the extent of melano- phore expansion is only slight stellation, yet this is a positive reaction. On examination, it was noticed that the stubs of the eye-stalks in those animals showing diurnal rhythm had healed, and were somewhat elongated. However, the stubs were scraped, cut off, or pulled out of their sockets without altering the slightly dark coloration, and conse- quently the possibility that the regeneration of the sinus gland had taken place was eliminated. That the slight darkening observed in blinded animals during daylight may be due to the direct action of light is improbable because this phenomenon is limited to only a few animals and occurs only after some time has elapsed following ex- tirpation.
Immersion of Animals in Distilled Water
Fifteen blinded animals were placed in a large dish containing dis- tilled water to a depth of one inch, which was changed daily. After 2 days of this treatment, 9 of the animals became dark (ranging from intermediate to complete darkness) while 6 remained pale. These were segregated, and listed as A and B groups respectively. On the third day, 4 of the animals in A were placed in sea water and designated as C, and the 3 groups watched daily:
|
A (5 animals) |
K (6 animals) |
C (4 animals) |
|
|
4th day |
All (lark |
1 pale, 5 dark |
All pale |
|
5th clay |
4 dark |
1 pale, 5 dark |
All pale |
|
6th day |
1 pale 3 dark |
2 pale, 4 dark |
All pale |
|
2 pale |
CHROA1ATOPHOROTROPIC HORMONE OF UCA
At the end of the week, one-third of all the animals had died, tin- remaining ten were placed in one dish, containing distilled water. On the eighth day, 6 had become pale, and on the ninth day all had become pale, remaining in this state for 4 more days when all but 2 died. One was now injected with distilled water, which promptly evoked an expansion within 15 minutes, but this response vanished within the next 15 minutes. The other was injected with 0.1 E.S. and responded in 15 minutes, remaining dark for 2 hours.
Distilled water has then the same melanophoric effect when ani- mals are immersed in it as when it is injected into the body spaces. After a week, however, the animals become refractory and the dark coloration is gradually lost, but the animals have not lost the ability to respond to injections of E.S.H. or distilled water.
Chemical Characteristics of E.S.H.
The properties of E.S.H. reported earlier (5) have been confirmed and extended. However, in our experiments, the solubility of E.S.H. in various solvents was determined after the following preliminary treatment: 100 eye-stalks were extracted three times with 10 cc. of distilled water by boiling. The solution was filtered and the filtrate reduced to a volume of 10 cc. under a warm current of air. Two volumes of either pyridine, ethanol or acetone were added, and the solution cooled to 5° C. for one day. A heavy red precipitate forms which is discarded after centrifugation. The supernatant fluid con- tains practically all of the activity, and after drying, was used as stock material. One extraction with 10 cc. of the following solvents was made of the dry powder equivalent to 10 E.S., and both soluble and insoluble fractions assayed:
Percentage of Solvent Activity Soluble
per cent
Absolute ethyl alcohol 45
Absolute methyl alcohol 60
95% ethyl alcohol 60
Absolute acetone 0
Ethyl ether 0
Petroleum ether 1
Benzene 1
Chloroform 0
Ethyl acetate 0
Pyridine 6
95% acetone 2
90% acetone 20
95% methyl alcohol 80
Due to the small amount of material available, we have not con- centrated on the purification of the hormone. However, if the stock
A. A. ABRAMOWITZ AXI) R. K. ABRAMOWITZ
material is repeatedly extracted with 95 per cent methyl alcohol, most of the activity can be collected as a soluble fraction, which, on basis of activity per milligram of dry weight, represents a tenfold purification. ;! Aqueous acetone seems to destroy some of the hormone for we have not been able to account for all of the original activity present before fractionation by totalling the activity of the soluble and insoluble fractions.
The hormone is apparently soluble in water throughout the pH range. Stock material (20 E.S.) was easily dissolved in 10 cc. of 0.1 N NaOH as well as in 10 cc. of 0.1 N HC'l with the formation of a floccu- lent, inactive precipitate. If 10 E.S. are dissolved in 10 cc. of 0.1
TABI.K III Diurnal rhythm of Uca pugilator
|
Time of Observation |
Normal Day ami NigM Conditions |
Constant Darkness |
||||
|
Normal Animals |
\niinals Blinded 2 Weeks Previously * |
Normal Animals |
Blinded 2 Weeks Previously |
|||
|
Indeterminate Background |
||||||
|
White Background 40 Animals |
Black Background 40 Animals |
Indeterminate Background 40 Animals |
Indeterminate Background 40 Animals |
|||
|
25 Animals |
25 Animals |
|||||
|
September 13 11:00 A.M. |
100% D |
100% D |
100% D |
7.V , 1-1' |
100%. D |
4%D |
|
7:30 P.M. |
82', I1 |
100% D |
100% D |
100% P |
— |
— |
|
12:45 A.M. |
100% P |
50% I; 50% P |
80% P |
100% P |
50% P; 50% I |
100% P |
|
September 14 10:00 A.M. |
100% D |
100% D |
100% P |
70% I-P |
100% D |
100% P |
|
8:00 I'M. |
26' •V 1 |
12% P; 88% P-I |
80% P |
08% P |
60% P |
100% P |
|
80% P; 20% I |
18% P: 82% I-D |
100% P |
98% P |
80% P |
100% P |
|
|
September 15 10:00 A.M. |
|i»r, |) |
100% D |
1110% D |
50% I-D |
100% D |
100% P |
|
4:30 P.M. |
100', D |
98% D |
100% D |
87% P |
25% P |
100% P |
|
10:30 P M. |
00% P |
50% P |
66% P |
87% P |
92% P |
100% P |
|
September IB 11:45 A.M. |
100' "0 D |
100% D |
100% D |
40% I |
96% D |
100% P |
|
10:30 P.M. |
100% P |
25% P 100% D |
100% P |
90% P |
84% P |
100% P |
|
September 17 10:30 A.M. |
100% D |
100% D |
40% I |
84% D |
100% P |
3 It was reported (8) that the hormone was precipitated from 95 per cent acetone by the addition of el her. This is an error since in this case the hormone was precipi- tated from 95 per cent nicthanol, not 95 per cent acetone.
CHROMATOPHOROTROPIC HORMONE OF UCA TABLE III — Continued
291
|
Time of Observation |
All Animals Now Exposed to Constant Light (75 Watt Electric Bulb) |
|||||
|
Normal Animals |
Animals Blinded 2 Weeks Previously * |
Normal Animals |
Blinded 2 Weeks Previously |
|||
|
Indeterminate Background |
||||||
|
White Background 40 Animals |
Black Background 40 Animals |
Indeterminate Background 40 Animals |
Indeterminate Background 40 Animals |
|||
|
25 Animals |
25 Animals |
|||||
|
September 18 11:15 A.M. |
100% D |
100'; D |
100% D |
40% I |
100% D |
55% I-P |
|
3:00 P.M. |
98% D |
100% D |
100% D |
25% I |
100% D |
50% I-P |
|
10:15 P.M. |
40% P; GO% I-D |
14% P |
33% P |
3%I;97%P |
33% P |
100% P |
|
September 19 12:30 P.M. |
100% D |
100% D |
100% D |
25% I |
100% D |
25% I-P |
|
10:00 P.M. |
30% P |
20%, P |
50% P |
92% P |
50% P |
100% P |
|
September 20 12:30 P.M. |
100%, D |
100%;, D |
100% D |
12% I |
100% D |
25% I-P |
|
11:00 P.M. |
75% P |
33% P |
75% P |
92% P |
50% P |
75% P |
|
September 21 12:30 P.M. |
100% D |
100%, D |
100% D |
14% I |
100% D |
50% I |
|
10:30 P.M. |
50% P |
20% P |
50% P |
90% P |
33% P |
90% P |
|
September 22 12:40 P.M. |
100% D |
100% D |
100% D |
16% I |
100% D |
42% I |
|
10:20 P.M. |
50% P |
16% P |
60% P |
94% P |
50% P |
90% P |
|
September 23 11:10 P.M. |
50% P |
6%P |
33% P |
91% P |
33% P |
100% P |
|
September 25 1:00 A.M. |
50% P |
14% P |
80%,, P |
96% P |
12% P |
75% P |
P = pale. D = dark. I = intermediate.
I-D = intermediate to dark. I-P = intermediate to pale.
* These were chosen from several hundred specimens because they were not entirely pale.
NaOH and, after 3 minutes, neutralized to pH 7.0 and tested, com- plete activity is found. If a similar solution is placed at 100° C. for ten minutes, total loss of activity occurs. Ten E.S. in 10 cc. of 0.1 N HC1 placed at 100° C. for ^ hour, neutralized, and tested, retains its original activity. In fact, HC1 seems to potentiate the hormone some- what, although we have not studied this carefully. Attempts to regenerate the activity of stock material treated for 15 minutes with NaOH by boiling it with an equivalent amount of HC1 after neutraliza- tion were unsuccessful.
292 A. A. ABRAMOWITZ AND R. K. ABRAMOWITZ
The Diurnal Rhythm of Uca pugilator
It has been stated (5) that the diurnal chromatic rhythm of Uca pugilator proceeded under constant illumination or darkness, and re- gardless of background. This statement was based upon observations made on only the melanophores of the abdominal segments, since the ventral surface of these crabs can be most easily observed. Since all the experiments reported here deal with the reactions of leg melano- phores, we studied for 2 weeks the behavior of the leg melanophores during the periodic change in coloration. Essentially, the same situa- tion as previously described was found, but several new facts of interest were discovered. The following protocol summarizes the observations upon 170 specimens kept under various conditions:
1 1 will be seen that when normal animals are exposed to normal day and night environment, the diurnal rhythm proceeds as described previously on both white and indeterminate backgrounds. On black background, however, the appearance of the nocturnal hue is delayed and does not occur in a certain percentage of animals. Under constant light, the same situation is true except that fewer animals on all back- grounds show the nocturnal phase. Animals maintained in darkness show the nocturnal coloration at night but when exposed to constant light, the number of specimens becoming pale at night is decreased. Two new facts therefore emerge: (1) that black background delays and inhibits the pale phase more effectively under conditions of constant illumination than under normal condition daylight and darkness, (2) that constant light does the same regardless of background.
Observations on the ventral surface of these specimens were made simultaneously but excluded from the protocol. If readings were based upon the behavior of the abdominal melanophores alone, the protocol would indicate that the pale phase occurred in practically all of the specimens regardless of background or of conditions of illumina- tion. There is, therefore, a decided difference between the reactions of the leg and abdominal melanophores.
The diurnal rhythm of Uca has been explained on the assumptions that during the day, K.S.H. is released into the blood stream and that during night, release is stopped. On this basis, it must be assumed that the abdominal melanophores are more sensitive than those of the legs to the disappearance of the hormone from the blood stream. In addition to the circumstantial evidence obtained by watching single animals or groups of animals during the diurnal rhythm, further proof was obtained by removing the eye-stalks of 15 dark animals during the day. After 1 hour, 20 per cent of the animals showed white
CHROMATOPHOROTROPIC HORMONE OF UCA
abdomens but the legs remained dark in all cases. After 1J/2 hours, 66 per cent of the specimens showed white abdomens and 20 per cent, pale legs. After 2}/2 hours, all showed pale abdomens while 70 per cent showed pale legs. This experiment, therefore, indicates quite clearly that the abdominal melanophores respond more rapidly than leg melanophores to the loss of E.S.H. from the blood, as brought about by removal of the eye-stalks.
The assumption that the pale phase of Uca is due to the absence of E.S.H. from the circulation during the night is based on no direct evidence inasmuch as we have not succeeded in obtaining the blood of Uca for assaying E.S.H. The evidence for this belief is, therefore, indirect, and is based on one fact — that pallor in Uca is always asso- ciated with and due to the loss of the eye-stalks, and hence removal of the chief if not the only significant source of E.S.H. However, it may be possible that the melanophores become refractory during the night to the eye-stalk hormone (which may be thought of as being secreted constantly) thus ushering in the pale phase, or that the pale phase is due to a melanophore-contracting hormone, concentrating nerves, or some unknown factor. The first possibility was negated by injecting some E.S.H. into 20 normal animals showing the nocturnal phase. All responded normally by darkening. The second possibility has no experimental basis inasmuch as no extract of Uca has been prepared which induces pallor, and as chromatophore nerves have not been demonstrated in the Crustacea.
Consequently, the explanation of the diurnal rhythm of Uca on the basis of the presence and absence in the blood stream of one hor- mone (E.S.H.) seems to be the most economical, and in conflict with no observational and experimental facts. The effect of constant light and of black background can therefore be thought of as prolonging the release of the eye-stalk hormone into the circulation, or as inhibiting or delaying the process by which the release of the hormone from the gland is stopped.
DISCUSSION
Since most of the data have already been analyzed, we would limit this to a short theoretical discussion on the nature of action of the eye-stalk hormone. It is evident that the problem of the endocrine control of chromatophores in crustaceans is not as simple as has been regarded. While in general the comparative endocrinology of melano- phore responses in both vertebrates and invertebrates is rather well understood, we have no inkling of the microphysiology of either the melanophores or the chromatophorotropic hormones in action. In
294 A. A. ABRAMOWITZ AND R. K. ABRAMOWITZ
fact, there is nothing known, with one or two exceptions, concerning the intimate mechanism by which a hormone is able to affect a tissue. The finding that distilled water expands the melanophores of blinded crabs may furnish a clue as to the manner in which the eye-stalk hor- mone acts, even though the reactions may not be comparable. Since the characteristic action of E.S.H. is the induction of protoplasmic movements in the melanophores, one is inclined to suspect the ionic or osmotic environment as the regulator of melanophore movements. This may or may not be reasonable but, at least, the effect of distilled water can be attributed to the resulting hypotonicity of the blood which in turn might cause a rapid shift in the osmotic equilibrium between the blood and tissues. The expansion of the melanophores might then result from such a change in the osmotic or ionic environment.
There are two possibilities concerning the way in which E.S.H. acts on the melanophores — directly or indirectly. Discussing the second possibility first, it is quite possible that the eye-stalk hormone may not have any relation to changeable coloration at all, but that it regulates the salt or water balance of the animal. The effect on melanophores may well be indirect, resulting from osmotic or ionic changes between the blood and tissues when the hormone is normal ly present and absent in the blood stream. This would mean, of course, that the chromatophorotropic hormone is only secondarily chromato- phorotropic; its main function being in regulating some other process. Indeed, it has been maintained (11) that since this hormone is found in crustaceans which do not show color changes or even contain chromato- phores, it must perform some function more important than that of regulating color changes. Such an argument is helpful for the above discussion, but we feel that the presence of a particular hormone may not necessarily mean that it must be performing some function. For example, intermedin is abundantly present in the pituitary glands of both birds and mammals, yet its presence is not correlated with any functional significance. Similarly, the presence of estrogenic hormones in the mammalian testes does not seem to be correlated with any of the normal processes in the reproductive physiology of the male. There are other instances of this kind, but of course some function for (lie hormones mentioned may be discovered eventually.
To the effect that E.S.H. acts directly on the melanophores, we might add that all of our curves relating response to concentration and time are of a type which has been assumed to represent reversible compound formation on cell surfaces — but this is all that we can marshal in support of a direct action. Isolated leg melanophores will expand when immersed into a solution of sea water containing E.S.H.,
CHROMATOP1IOROTROPIC HORMONE OF UCA
hut this is of course no indication of a direct action. Similarly, the effectiveness of various salts on isolated melanophores offers nothing to indicate that the action of E.S.H. is direct or indirect. The deter- mination of the mechanism of action of the eye-stalk hormone belongs properly to general physiology and is perhaps one of the most signifi- cant aspects of melanophore physiology. It is hoped that some of these observations will furnish a basis for further investigation.
In concluding, it can be mentioned that a search for other endocrine functions of the eye-stalk such as have already been found (regulation of heart (12), calcium metabolism (13)) will be of importance in establishing the proper role which the eye-stalks play in regulating the pigmentary effectors. The hormone is not indispensible to the life of crustaceans. Blinded crabs molt frequently (perhaps more frequently than normal specimens according to our impression) and if well fed can be maintained in a healthy state for several months, and perhaps indefinitely.
SUMMARY
1. A method for determining within an hour the relative activity of eye-stalk extracts of Uca pugilator has been described.
2. The amount of hormone present in one eye-stalk of Uca (5.0 grams) was calculated to be 0.2 7. The minimal amount of hormone detected by the method described was calculated to be 0.000016 7, which represents a minimal unit of eye-stalk activity.
3. Distilled wrater is effective in expanding melanophores when injected into blinded specimens. The effect seems to be related to osmotic changes induced in the animal. Excepting the effect of dis- tilled and tap water, the response of blinded animals to the eye-stalk hormone seems to be quite specific. Sixteen different drugs were ineffective in causing melanophore expansion.
4. Blinded animals immersed in distilled water become and remain dark for a week, after which they become refractory. However, they have not lost the ability to respond to injections of either eye-stalk hormone or distilled water.
5. Extracts of several organs other than the eye-stalk were found to be inactive on blinded specimens, even when 100 times more con- centrated than a minimal unit of eye-stalk extract. Extracts of entire bodies, 2,000 times more concentrated than a minimal unit, were active.
6. Isolated leg melanophores respond slowly and irregularly to various ions. Na, K, Ca, Li chlorides all induce melanophore con- traction. Na and K do not produce opposed responses in isolated leg melanophores as they do in isolated scale melanophores.
296 A. A. ABRAMOWITZ AND R. K. ABRAMOWITZ
7. The relative solubility of the eye-stalk hormone in various organic solvents was determined. The hormone is inactivated by alkali, but the activity cannot be regenerated by treatment with acid.
8. Animals blinded for 2 weeks or more remain continuously pale, with the exception of 2-4 per cent of the specimens, which show slight periodic changes in coloration.
0. The diurnal rhythm of Uca is affected by constant light, and black background, both of which delay the appearance of the nocturnal hue on the lei;-.
LITERATURE CITED
1. Mi :t,\>i K. I-"., 1912. Arch.f. Entw.-mech., 33: 462.
2. AKRAMOUII/. A A., 1935. Proc. Nat. Acad. Sci., 21: 677.
3. CARLSON, S. I'.. 1935. Proc. Nat. Acad. Sci., 21: 549.
4. ABKAMOWITZ, A. A., 1938. Phvsiol. Zonl. fin press).
5. AHKAMOWITZ, A. A., 1937. Biol. Hull.. 72: 344.
6. HANSTROM, B., 1931. Zeitsclir. f. Morph. u. Okol. d. Ticrc, 23: 80.
7. AHRAMOWITZ, A. A., 1936. Proc. Nut. Acad. Sci., 22: 521.
8. AHRAMOWITZ, A. A., 1936. Anal. Rcc. (Supplement No. 1) (Abstract) 67: 108.
9. BROWN, !•'. A., JR., 1935. J,,nr. I-'.xper. /.».•>/., 71: 1. ID. SPAETH, K. A., 1913. Jour. l-:.\[)cr. Z<>«1., 15: 527.
11. HANSTROM, B., 1937. Kungl. Svcnska Vctcn. Hand., 16: 3.
12. WELSH, J. H., 1937. Proc. Nat. Acad. Sci.. 23: 458.
13. KOLI.KR, G., 1930. Zeitschr. ver»l. PhysioL, 12: 632.
THE EFFECT OF SODIUM, POTASSIUM AND CALCIUM
IONS ON CHANGES IN VOLUME OF
AMOEBA PROTEUS1
S. O. MAST AND COLEEN FOWLER
(From the Zoological Laboratory of the Johns Hopkins University)
INTRODUCTION
McCutcheon and Lucke (1928) found that, with the same osmotic concentration, permeability of the eggs of Arbacia to water is greater in solutions of dextrose containing 0.01 M NaCl 6r KC1 and less in solutions of dextrose containing 0.004 M CaCl2 or MgCl2 than in solutions of dextrose in distilled water. These results indicate that sodium and potassium ions increase and calcium and magnesium ions decrease permeability to water. Jacobs (1930) showed that hemolysis of erythrocytes in solutions of non-electrolytes, 0.01 M or 0.02 M, is almost as rapid as it is in distilled water; but that in solutions of electrolytes, e.g. NaCl 0.001 M or CaCl2 0.0001 M, it is considerably slower. These results indicate that sodium as well as calcium ions decrease permeability to water, i.e. that under some conditions mono- valent as well as divalent salts cause decrease in permeability.
The experiments considered in the following pages were under- taken to ascertain the effect of sodium, potassium and calcium ions in phosphate buffer solutions (0.002 M, pH 6.8) on changes in the volume of Amoeba proteus.
MATERIALS
Amoeba proteus was used exclusively in these experiments. It was grown in cultures prepared according to the method described in an earlier paper (Mast and Fowler, 1935). The distilled water used was triple redistilled from a tandem glass still (Mast, 1928). The solutions used were prepared by making two stocks, one containing the primary phosphate of the cation to be tested, 0.002 M, and the other containing the hydroxide of the same cation, 0.002 M, and mixing the two stocks
1 This paper is the second in a series on the relation between environmental factors and water-content in Amoeba proteus. The first paper in the series, entitled " Permeability of Amoeba proteus to Water," was published in the Journal of Cellular and Comparative Physiology, Vol. 6, pp. 151-167, 1935.
These investigations were supported by a grant received by the junior author from the Brooks Fund of the Zoological Laboratory of the Johns Hopkins University.
297
298
S. O. MAST AND COLEEX FOWLER
in the proportions necessary to maintain the hydrogen ion concentra- tion at pH 6.8. The preparation and standardization of the hydroxide stock solutions is described in the paper cited above. All the solutions thus prepared contained the same concentration of the metallic cation used and the osmotic concentration of these solutions was practically (he same as that in which the amoebae were grown, i.e. modified Ringer solution.
If the salts in the solutions used had been completely ionized, the osmotic concentration of the solutions containing calcium would have been greater than I hose containing either sodium or potassium because if CaH.i(PO.i)2 ionizes there are produced two phosphate ions for each
DISTILLED WATER
CALCIUM SOLUTION 0.002 M
SODIUM SOLUTION O.002M
POTASSIUM SOLUTION
15
30 45 TIME IN MINUTES
FIG. 1. Graphs showing the relation between sodium, potassium and calcium respectively in phosphate buffer solutions (pH 6.8) and changes in volume in Amoeba protens transferred from dilute Ringer solution of the same osmotic concentration. Each point on the curves represents the average of the results of three measurements made on three different individuals. Ordinates, change in volume as percentage of the original volume; abscissae, time in minutes after transfer from dilute Ringer solution.
calcium ion, when M- il KILPOi or XaH.>P()4 ionizes there is produced only one phosphate ion for each potassium or sodium ion. At the concentration used, however, the sodium and the potassium com- pounds were slightly more ionized than the calcium compounds. The number of ions per unit volume and the osmotic pressure were probably therefore practically the same in the solutions used.
METHODS AND RESULTS
Three methods were used in ascertaining the effect of these solutions on changes in volume of Amoeba.
EFFECTS OF SALTS ON VOLUME CHANGES IN AMOEBA 299
1. Six amoebae were prepared, selected, and measured individually in modified Ringer solution as described in the paper cited above (Mast and Fowler, 1935). Then they were transferred to distilled water and measured at fifteen-minute intervals. Then this was repeated with other individuals, using in place of distilled water phosphate buffer solutions which contained sodium, potassium and calcium respectively, but with only three individuals in each. The results obtained are presented in Fig. 1.
HUM SOLUTION &002M
15
JO 45 60 TIME IN MINUTES
FIG. 2. Graphs showing the relation between sodium, potassium and calcium respectively in hypertonic solutions of lactose (pH 6.8) and changes in volume in Amoeba proteus transferred from dilute Ringer solution.
Each point on the curves represents the average of the results of ten measure- ments made on 7 to 12 different individuals. Ordinates, change in volume as per- centage of the original volume; abscissae, time in minutes after transfer.
Figure 1 indicates that the amoebae transferred from modified Ringer solution to distilled water increased 10 per cent in volume in 75 minutes; that those transferred to the calcium phosphate buffer solutions of practically the same osmotic concentration as the Ringer solution, did not change appreciably in volume during the time re- presented in the graph, and that those transferred to the sodium phosphate buffer solution decreased only slightly if at all during this time, but that those transferred to the potassium phosphate buffer
300
S. O. MAST AND COLEEX FOWLER
solution of the same osmotic concentration as the preceding solutions, decreased 20 per cent in 60 minutes and then remained constant.
After the experiment had continued 105 minutes, the amoebae in the distilled water and in the calcium phosphate buffer solution were normal and they continued to live for several days; those in the sodium phosphate buffer solution soon began to round up and decrease in volume and they disintegrated about 60 minutes later; and those in the potassium buffer solutions disintegrated even sooner. However, those in all the solutions recovered fully in culture solution, if they had not been in the salt solutions more than one hour. The results obtained show, therefore, that without any change in osmotic concen-
TABLE I
Percentage changes in volume in amoebae transferred from modified Ringer solu- tion (0.002 M, pH 6.8) to a solution containing sodium as the only metallic cation, of the same hydrogen ion and osmotic concentration as the Ringer solution, plus 0.2 M lactose. Temperature, 22.7°-25.5°; £,„, probable error.
|
Designation of specimen |
Time in minutes after transfer |
||||||
|
15 |
30 |
45 |
60 |
75 |
90 |
105 |
|
|
1 |
- 7 |
-17 |
-17 |
-27 |
-32 |
-32 |
-46 |
|
2 |
- 2 |
-18 |
-37 |
-48 |
-55 |
-60 |
-55 |
|
3 |
-13 |
-20 |
-25 |
-29 |
-34 |
-46 |
-46 |
|
4 |
-14 |
-23 |
-26 |
-40 |
-53 |
-52 |
-62 |
|
5 |
- 8 |
-24 |
-45 |
-61 |
-63 |
||
|
6 |
-17 |
-20 |
-28 |
-26 |
-33 |
-35 |
-43 |
|
7 |
+ 1 |
- 9 |
-21 |
-31 |
-36 |
-41 |
-44 |
|
8 |
-16 |
9 |
j |
-41 |
-36 |
-27 |
-27 |
|
9 |
-18 |
-26 |
-23 |
-34 |
-41 |
-39 |
-51 |
|
10 |
- 3 |
- 5 |
- 8 |
-16 |
-41 |
-58 |
|
|
Mean |
-10 |
-17 |
-23 |
-35 |
-42 |
-43 |
-47 |
|
Em |
±1.382 |
± 1 .442 |
±2.36 |
±2.58 |
±2.18 |
±2.44 |
±2.28 |
tration potassium causes rapid loss of water in Amoeba, without any appreciable injury; and that sodium and calcium have little if any effect.
2. Several amoebae were prepared, selected, and measured as in method 1, and then transferred to 0.2 M lactose in distilled water, and measured at fifteen-minute intervals. Then this was repeated \\itli phosphate buffer solutions which contained lactose (0.2 M) and sodium, potassium and calcium (0.002 AI) respectively. The results obtained are presented in Fig. 2 and Tables I, II, and III.
Figure 2 indicates that the amoebae which were transferred from modified Ringer solution to hypertonic solutions, consisting of lactose
EFFECTS OF SALTS ON VOLUME CHANGES IN AMOEBA 301
in distilled water and respectively sodium, potassium and calcium plus lactose, decreased in volume in all the solutions, but that the decrease was least in the solution containing calcium and greatest in distilled water plus lactose.
Tables I, II, and III show that there was marked individual varia- tion in the reduction in the volume of the amoebae in all the hypertonic solutions tested, but that in spite of this variation, the probable error is relatively so small that there is overlapping at only a few points on the curves. This practically proves that the differences in the means of the reduction in volume in the different solutions, indicated in the
TABLE II
Percentage changes in volume in amoebae transferred from modified Ringer solutions (0.002 M, pH 6.8) to a solution containing potassium as the only metallic cation, of the same hydrogen ion and osmotic concentration as the Ringer solution, plus 0.2 M lactose. Temperature, 22.1°-25.2°; £„„ probable error.
|
Designation of specimen |
Time in minutes after transfer |
||||||
|
IS |
30 |
45 |
60 |
75 |
90 |
105 |
|
|
1 |
- 2 |
- 8 |
-19 |
-23 |
-28 |
-23 |
-36 |
|
2 |
-10 |
-21 |
-24 |
-30 |
-28 |
-35 |
-33 |
|
3 |
+ 10 |
-11 |
-16 |
-23 |
-35 |
-43 |
|
|
4 |
- 2 |
-10 |
-15 |
-15 |
-28 |
-20 |
-34 |
|
5 |
-21 |
-14 |
-22 |
-31 |
-24 |
-43 |
-43 |
|
6 |
-11 |
-17 |
-23 |
-32 |
-42 |
-57 |
|
|
7 |
-11 |
-14 |
-11 |
-30 |
-37 |
-43 |
|
|
8 |
+ 10 |
- 4 |
|||||
|
9 |
- 4 |
- 4 |
-11 |
-29 |
-37 |
||
|
10 |
-17 |
-17 |
-17 |
-20 |
-20 |
-46 |
|
|
11 |
- 9 |
- 9 |
-16 |
-20 |
-34 |
-43 |
|
|
12 |
-11 |
-11 |
-18 |
-24 |
-32 |
-42 |
-58 |
|
Mean |
• 6 |
-12 |
-18 |
-25 |
-32 |
-40 |
-41 |
|
•C*ni |
±1.79 |
±0.97 |
±0.85 |
±1.07 |
±1.26 |
±2.10 |
±2.80 |
curves in Fig. 2, are significant and consequently that the order of effectiveness on preventing loss of water in Amoeba, is Ca>K>Na, with very little difference between calcium and potassium.
The facts that the osmotic concentrations of the solutions which contained respectively sodium, potassium and calcium plus lactose were practically equal and that the osmotic concentration of these solutions was somewhat higher than that of the distilled water plus lactose, show that the monovalent sodium and potassium ions as well as the divalent calcium ions, retard the loss of water, but that calcium ions are more effective than sodium or potassium ions.
302
S. O. MAST AND COLEEN FOWLER
3. Ten amoebae were prepared, selected, and measured in modified Ringer solution as in methods 1 and 2, then they were transferred to this solution containing lactose (0.2 M), left 30 minutes and measured again. They were then transferred to distilled water and measured at 30-minute intervals. This whole process was then repeated with other specimens in respectively sodium, potassium and calcium phosphate buffer solutions (0.002 M) in place of distilled water. The results obtained are presented in Fig. 3.
Figure 3 shows that the amoebae which were transferred from modified Ringer solution to modified Ringer solution containing lactose (0.2 M) decrea>e<l in volume to about 70 percent of the original volume in 30 minutes and that after they had been transferred from this solution to distilled water or to phosphate buffer solutions containing
TABLE III
Percentage changes in volume in amoebae transferred from modified Ringer solu- tion (0.002 M, pH 6.8) to a solution containing calcium as the only metallic cation, of the same hydrogen ion and osmotic concentration as the Ringer solution, plus 0.2 M lactose. Temperature, 22.5°-24.0°; Em, probable error.
|
Designation of specimen |
Time in minutes after transfer |
||||||
|
IS |
30 |
45 |
60 |
75 |
90 |
105 |
|
|
1 |
- 8 |
- 2 |
-11 |
-20 |
-23 |
-29 |
|
|
2 |
-12 |
-22 |
-25 |
-17 |
-29 |
-44 |
-53 |
|
3 |
-12 |
-21 |
-23 |
-26 |
-41 |
||
|
4 |
+ 7 |
- 8 |
• 6 |
-13 |
-30 |
-36 |
-43 |
|
5 |
+ 4 |
- 5 |
- 8 |
||||
|
6 |
+ 7 |
-18 |
-27 |
-29 |
-32 |
-38 |
-47 |
|
7 |
- 3 |
- 5 |
-21 |
-17 |
-13 |
-42 |
|
|
Mean |
- 2 |
-12 |
-17 |
-21 |
-28 |
-38 |
-48 |
|
Em |
±2.01 |
±1.98 |
±2.01 |
± 1 .53 |
±1.93 |
±1.55 |
±1.61 |
sodium, potassium or calcium, they increased in volume, but that the increa>r was much more rapid and much greater in distilled water than in any of the salt solutions. The figure shows that in all tin- salt solutions the amoebae increased to a maximum and then decreased, bin that increase was greater in the solution which contained sodium than those which contained potassium or calcium, the maximum increase being 22 per cent for sodium and only ^ per cent for potassium and 4 percent for calcium. It also shows that in distilled water the amoebae nearly regained their original volume in 30 minutes, and then continued to increase as long as measurements were made, i.e. for <>() minutes, and that at the close of the experiment they were 17 per cent larger than at the beginning.
EFFECTS OF SALTS ON VOLUME CHANGES IN AMOEBA 303
The individual variations in the increase in volume in the hypotonic solutions were essentially the same as those recorded in Tables I, II and III for hypertonic solutions, but the differences in the means were greater so that there was no overlapping of the probable errors, except at the last points on the curves. The differences in the means of the increase in volume in the different solutions indicated in the curves in Fig. 3, are therefore significant, showing that the order of
30 30
TIME IN MINUTES
FIG. 3. Graphs showing the relation between sodium, potassium and calcium respectively in hypotonic buffer solutions (pH 6.8) and changes in volume in Amoeba proteus transferred from weak Ringer solution containing 0.2 M lactose.
Each point on the curves represents the average of the results obtained in ten measurements made on ten different individuals. Ordinates, change in volume as percentage of the original volume; abscissae, time in minutes after transfer.
effectiveness on gain of water in Amoeba in hypotonic solutions is, Ca>K>Na, that is, that it is the same as it is on loss of water in hypertonic solutions.
By comparing Fig. 3 with Fig. 1, it will be seen that if Amoeba is transferred from Ringer solution (0.002 M) to potassium phosphate solution (0.002 M) it decreases markedly in volume, but that if it is transferred from lactose solution (0.2 M) to potassium phosphate
304 S. O. MAST AXU COLEEX FOWLER
solution (0.002 M) it increases somewhat. This difference is doubtless correlated with the fact that the water-content was much higher in the amoebae which were transferred from the modified Ringer solution than it was in those which were transferred from the lactose solution. The results presented in Fig. 3 show therefore that monovalent as well as divalent cations retard the entrance of water but that, just as in reference to loss of water, the latter are more effective than the former; and they support the conclusion reached above that both monovalent and divalent metallic cations tend to maintain the water- content in Amoeba constant.
DISCUSSION
The question now arises as to how the metallic ions act in their tendency to maintain the water-content of Amoeba constant.
Under normal conditions much water is eliminated by the contrac- tile vacuole and considerable water enters with the food in the process of feeding (Mast and Hahnert, 1935) and some possibly enters in the formation of cups without food (Mast and Doyle, 1934). It may be then that the metallic ions influence the water-content of Amoeba through action on some or all of these processes. Such action could, however, not have been seriously involved in the results presented in the preceding pages; for in the solutions used there was, during the time the observations were made, no feeding and no formation of cups of any sort and the contractile vacuole did not function except in a few of the more dilute solutions and then only after the amoebae had been in the solution half an hour or more. The changes in water-content observed must therefore have been due to passage of water directly through the surface layer of the amoebae, and the effects of the metallic ions on the rate of the passage through this layer must have been due either to their action on the permeability of this layer or to their action on the osmotic concentration of substances in the amoebae, or to their action on imbibition, but it is probably largely due to the first. This matter will be discussed more fully in a subsequent paper.
The conclusion that monovalent as well as divalent metallic cations tend to maintain the w.ater-content of Amoeba constant, is in agreement with the results obtained by Jacobs, who showed that in erythrocytes permeability to water is decreased by the addition of eit her Na( 'I or ( "a( 'I- to (list illed water; but they are not in agreement with those obtained by McCutcheon and Lucke, who showed that permeability of the eggs of Arbacia to water is increased if NaCl or KC1 is added to a solution ol dextrose in distilled water and decreased if CaCl2 or MgCl2 is added.
EFFECTS OF SALTS ON YOLl'MK CHANGES IN AMOEBA 305
SUMMARY
1. If amoebae are transferred from modified Ringer solution (0.002 M, pH 6.8) to distilled water or to phosphate buffer solution containing respectively sodium, potassium and calcium as the only metallic cation (0.002 M, pH 6.8), they increase in size in distilled water, decrease in size in the solution which contains potassium and remain nearly the same in size in the solution which contains sodium or calcium.
2. If they are transferred from modified Ringer solution (0.002 M, pH 6.8) to distilled water containing lactose (0.2 M) or to phosphate buffer solutions containing lactose (0.2 M) and respectively sodium, potassium and calcium as the only metallic cation, they decrease in volume in all the solutions but they decrease most rapidly in distilled water containing lactose, less rapidly in the solution containing sodium and lactose, still less rapidly in the solution containing potassium and lactose, and least rapidly in the solution containing calcium and lactose.
3. If they are transferred from modified Ringer solution (0.002 M, pH 6.8) to Ringer solution containing lactose (0.2 M) and left 30 minutes, they decrease greatly in volume. If they are then trans- ferred respectively to distilled water and to phosphate buffer solutions, containing respectively sodium, potassium and calcium (0.002 M, pH 6.8), they increase in volume in all the solutions, but they increase most rapidly in distilled water, less rapidly in the solution containing sodium, still less rapidly in the solution containing potassium and least rapidly in the solution containing calcium.
4. These results show that under some conditions monovalent as well as divalent cations cause decrease in the rate at which water passed into and out of Amoeba proteus. This is probably largely due to their action on permeability to water.
LITERATURE CITED
JACOBS, M. H., 1930. The influence of electrolytes on certain types of hemolysis.
Am. Jour. Med. Sci., 179: 302-303. MAST, S. O., 1928. Factors involved in changes in form in Amoeba. Jour. Expcr.
ZooL, 51: 97-120. MAST, S. O., AND W. L. DOYLE, 1934. Ingestion of fluid by Amoeba. Protoplasrna,
20: 555-560. MAST, S. O., AND COLEEN FOWLER, 1935. Permeability of Amoeba proteus to water.
Jour. Cell, and Comp. Physiol., 6: 151-167. MAST, S. O., AND W. F. HAHNERT, 1935. Feeding, digestion, and starvation in
Amoeba proteus (Leidy). Physiol. ZooL, 8: 255-272.
McCuTCHEON, M., AND B. LUCRE, 1928. The effect of certain electrolytes and non- electrolytes on permeability of living cells to water. Jour. Gen. Physiol.,
12: 129-138.
THE CONTRACTILE VACUOLE IN AMOEBA PROTEUS (LEIDV)
S. O. MAST
(From tlte Zoological Laboratory of the Johns Hopkins University and tlie Marine Biological Laboratory, Woods Hole, Mass.)
Metcalf (1910) maintains that there are scattered through the cytoplasm of Amoeba proteus numerous small round granules called microsomes or cytomicrosomes, and that the "contractile vacuole is surrounded by a layer of granules of the same size and appearance as the microsomes of the general cytoplasm." He says (p. 302): "When the vacuole is of moderate size these granules form a continuous layer, one granule thick," when it is large there are open spaces between the granules, and when it is small the layer is several granules thick. He maintains that during contraction these granules form a clump in which the new vacuole develops and consequently concludes that they "are connected functionally with excretion," and he therefore calls them "excretion granules." He admits, however, that the contractile vacuole can arise and function without the granules. He says (p. 307) ; "The results of operations upon Amoeba show that a new vacuole in an Amoeba fragment appears first and that only gradually the granules collect about it. The granules are not essential to the functional vacuole."
Mast (1926) confirmed Metcalf's contentions concerning the form, size and distribution of the granules in question, but he holds that the conclusion that they function in excretion is not well founded and that their aggregation on the surface of the contractile vacuole is a purely physical phenomenon. He designated them beta granules to differ- entiate them from somewhat similar but much smaller granules which he designated alpha granules.
MUM and Doyle (1935a, b] demonstrated that the beta granules have staining properties like the granules in other cells, known as mitochondria. They cent rifuged amoebae and found thai all the beta granules moved through the cytoplasm from the axis of rotation and that those on the contractile vacuole moved to the centrifugal surface and formed a considerable mass there, and they obtained some evi- dence indicating that the beta granules on the surface of the contractile vacuole are imbedded in a layer of substance which is somewhat more
306
CONTRACTILE YACUOLE IN AMOEBA PROTEUS (LEIDV) 307
viscous than the adjoining substance in the cytoplasm and moves with them during the process of centrifuging. They conclude that the beta granules function in transferring substances from place to place in the cytoplasm and that they probably thus facilitate excretion of substance by the contractile vacuole.
METHODS
I made at various times numerous detailed observations on the contractile vacuole in Amoeba proteus under the best optical system obtainable (oil immersion apochromatic objective, 60 X and 90 X, and compensating oculars, 10 X, 15 X and 20 X). All the observa- tions were made on living specimens. Some of the specimens had been without food for several days and contained no food vacuoles and but little fat and relatively few granules and crystals. Some of these were greatly flattened by gradually removing water from under the coverglass. Under these conditions the contractile vacuole and the granules and other substances around it could be very clearly seen. The following results were obtained.
RESULTS
The contractile vacuole is usually surrounded by a layer of beta granules (Metcalf's excretory granules). These granules are imbedded in a layer of substance which is optically distinctly different from the adjoining cytoplasm. It has a slightly yellowish tint and appears to consist of coagulated substance. It is more viscous and heavier than the rest of the cytoplasm for it moves to the centrifugal surface of the contractile vacuole when amoebae are centrifuged (Mast and Doyle, 19356, Fig. 1). When the contractile vacuole is fully developed the granules around it usually form a layer one granule thick, with some spaces between the granules here and there. As the vacuole contracts these spaces disappear, the layer of granules becomes continuous and the layer of substance in which they are imbedded thickens. As it contracts further this layer becomes still thicker and the granules pile up on each other until there is, when contraction is complete, a distinct mass of substance of considerable size, in which a remnant of the vacuole can usually still be seen. That is, the vacuole usually does not completely disappear during systole and soon begins to enlarge again. If it disappears completely a new vacuole usually appears and develops in the mass of substance in which the old one disappeared (Fig. 1). This, with the exception of the layer of substance in which the granules are imbedded, is in full accord with Metcalf's views (1910).
At the surface of the contractile vacuole under the layer of sub- stance containing the beta granules, there is a layer or membrane about
308
S. O. MAST
0.5/z thick which is optically well differentiated from the adjoining substance on either surface, for under favorable conditions a line indicating an interface can be clearly seen at both of these surfaces. The beta granules are never in close contact with this membrane.
ta-
Fi<r. 1. Camera drawings of the contractile vacuole in Amoeba protcus showing tlu- relation between the vacuole and the surrounding cytoplasm, under different conditions and in different phases. »;, membrane on the surface of the vacuole; a, layer of differentiated cytoplasm around the vacuole; b, beta granules (mitochondria) imbedded in this layer; /, plasmalrmma ; <;, pla-magel; .v, plasmasol; H, nucleus.
.1, (oni ractile vacuole partially expanded; II, same vacuole fully expanded; C, same vacuole shortly after contraction; D, a vacuole which had only a few beta granule;- in the dilferent iaied layer at the surface; E, F, C, contractile vacuoles which are in contact with other structure-.
Note that the beta granules are not in contact with the membrane at the surface of the vacuoles, that thev differ greatly in number in different vacuoles, and that the layer of differentiated < -\toplaMii in which they are imbedded becomes much thicker as the vacuole contract-. Note also that when the contractile vacuole comes in contact with other structure^ the differentiated layer around the vacuole is pushed aside.
They are always separated from ii by substance in the layer in which they are imbedded. This separation is least when the contractile vacuole is fully developed and increases as it contracts (Fig. 1 ).
II the contractile vacuole conies in contact with an object, the layer
CONTRACTILE VACUOLE IN AMOEBA PROTEUS (LEIDY) 300
in which the beta granules are imbedded is pushed aside and the membrane around it comes in direct contact with the rest of the cyto- plasm (Fig. 1). This shows that this layer is only loosely connected with the membrane.
The number of beta granules varies greatly and probably also the amount of substance in which they are imbedded. I have seen a considerable number of fully developed vacuoles which functioned normally in all respects, with only 10 or 12 granules at the surface (Fig. 1), and one in which there were no granules and no differentiated layer of substance. This indicates that these granules and the layer of substance in wyhich they are imbedded have but little if any func- tional relation with the contractile vacuole. The layer of substance is probably due to the action of the fluid in the vacuole on the adjoining cytoplasm.
DISCUSSION
There is considerable confusion in reference to the ideas held by different investigators concerning the designation of the structure and the function of the granules and crystals in the cytoplasm of the protozoa.
Butschli (1881, Abt. 1, S. 103) maintains that there are in fresh- water rhizopods and other protozoa, numerous highly refractive bodies ("Korner") which vary greatly in size "von ausserster Kleinheit bis zu ziemlich ansehnlichen Dimensionen," and frequently appear in the form of rhomboid crystals, and that they are not composed of fat. He holds that these bodies are end-products of metabolism and he calls them "Excretkornchen." He obviously, therefore, includes under this term various kinds of granules and crystals, that is, all which are highly refractive except those wrhich are composed of fat.
Schewiakoff (1893) found in Paramecium small bodies he called "Exkretkorner." He studied these bodies in considerable detail and concluded that they consist of calcium phosphate and are optically active, "doppeltbrechend im polarisirten Lichte." He says (p. 32) "In Folge der Doppeltbrechung, welche bereits von Maupas ermittelt wurde, sind sie von anderen Plasmaeinschliissen leicht zu unterschei- den." Schewiakoff' s Exkretkorner therefore include only a portion of Biitschli's Excretkornchen, in fact only a very small portion, for only a very small percentage of the granules and crystals in Amoeba proteus are optically active.
Metcalf (1910), as previously stated, maintains that there are on the surface of the contractile vacuole and scattered through the cyto- plasm of Amoeba proteus small round granules. He calls them " micro- somes," "Cytomicrosomes," or "excretory granules." Must and
310 S. O. MAST
Doyle (1935) demonstrated that these granules are not optically active. They are therefore not like the Exkretkorner of Schewiakoff.
Taylor (1923. p. 266) says that certain granules which appear throughout the cytoplasm of Enplotes are apparently comparable with Metcalf's "cytomicrosomes" in Amoeba and Opaline (1910), Biitschli's "Exkretkorner' (1881), Maupas' "corpuscle refringents" (1883), and Schcwiakoffs "Exkretkorner" in Paramecium.
Since Metcalf s cytomicrosomes differ radically from SchewiakofFs Exkretkorner, it is difficult to see how the granules in Euplotes can be comparable with both. Taylor does not say whether or not they are optically active but he calls them "crystaloidal, endoplasmic granules" and he says " I have never observed that they tend to aggregate about tin- contractile vacuole." They are then-fore probably more nearly like Schewiakoff s Exkretkorner than Metcalf's "cytomicrosomes" or "excretory granules."
Taylor (p. 278) maintains that Khainsky (1911) observed granules pass through the walls of the food vacuoles into the cytoplasm and he says: "These crystalloids he [Khainsky] identifies with the Exkret- korner described at length by Schewiakoff. Khainsky thinks that eventually the granules are dissolved in the endoplasm and the solution is discharged to the outside by the contractile vacuole." I find that Khainsky presents evidence which indicates that granules pass from the food vacuoles directly into the surrounding cytoplasm, but I find nothing which indicates that he identified these granules with Schewiakoffs Exkretkorner or that he held that they are dis- solved in the endoplasm and discharged through the contractile vacuole.
The fact that Biitschli designated the granules and crystals he observed " Excretko rnchen " clearly indicates that he held that they are involved in excretion. He says nothing, however, concerning this.
Schewiakoff maintains that the bodies he designated " Exkret- korner" dissolve in the cytoplasm and are excreted in fluid form by the contractile vacuole. He says (p. 55): " Es liegt demnach die Yermut- hung nahe, dass die Exkretkorner im Protoplasma aufgelost und ini fiissigen Zustande (lurch die kontraktile Vacuole nach ausen entleert werden."
Metcalf (1910) thinks that his "excretory granules" are "func- tionally connected with excretion." He gives nothing concerning the method involved, but his description of the relation between these granules and the action of the contractile vacuole shows clearly that he holds that they are not dissolved and discharged by the contractile vacuole. His conception concerning tin- method of function of these
CONTRACTILE VACUOLE IN AMOEBA PROTEUS (LEIDY) 311
granules consequently differs radically from Schewiakoff's concerning the method of function of his Exkretkorner.
Taylor (1923, p. 277) says: " Vacuoles (contractile) in Euplotes make their first appearance either from the coalescence of other extremely minute vacuoles that easily escape detection, or from the transforma- tion of small vacuoles which contain one or more granules ('Exkret- korner'), or they arise de novo." I can see no essential difference between the first possibility and the third.
Concerning the second possibilities, Taylor thinks the granules in question may pass from the food vacuoles into the cytoplasm in accord with Khainsky's observations referred to above, that vacuoles may thus form around these granules, that the granules may then dissolve, and that the fluid vacuoles thus formed may develop into contractile vacuoles; and he holds that if this obtains the origin of the contractile vacuole is traceable to the food vacuole. The fact, however, that the contractile vacuoles continue to operate normally for days in the total absence of food, that long after all vacuoles containing food and the granules which have passed from them into the cytoplasm have disappeared, practically proves that they do not originate in this way.
Concerning the third possibility Taylor says: "Now it is conceivable that here and there in this organism ' fluid centers ' normally arise which contain in solution substances (e.g. catabolic products) of a kind and concentration sufficient to induce the gelation of the surrounding plasm, thereby forming de novo normal vacuoles."
This possibility is in full accord with Metcalf's views concerning the origin of the contractile vacuole, except that he holds that the beta granules ("excretory granules") are involved. The fact, however, that under some conditions the contractile vacuole forms in regions where there are but few if any of these granules and that it develops normally at times when there are practically none on the surface of it, demonstrates that if they are at all functionally connected with the origin or the development of this vacuole, it is only in a very minor way.
It may be said then that while it is highly probable that the con- tractile vacuole originates in minute localized aggregations of fluid in the cytoplasm, nothing is known concerning the factors involved.
There is no evidence concerning the function of the relatively viscous layer in which the beta granules are imbedded on the surface of the contractile vacuole.
SUMMARY
1 . The contractile vacuole in Amoeba proteus contains at the surface a well differentiated membrane about 0.5/x thick. Adjoining this
312 S. O. MAST
membrane on the outside there is usually a layer of substance in which numerous beta granules (Metcalf's "excretory granules") are imbed- ded. This layer is more viscous and heavier than the adjoining cytoplasm. It is usually about 3yu thick when the vacuole is maximum in size and it becomes thicker during contraction.
2. The lid a granules around tin- contractile vacuole vary great ly in number and the layer of substance in which they are imbedded varies greatly in thickness, without any apparent variation in the function of the vacuole. These facts indicate that neither the granules nor the layer of substance is involved in the function of the contractile vacuole, at least not directly.
3. The differentiation of a layer of substance on the surface of the contractile vacuole is probably due to the action of the fluid in the \acuole on the adjoining cytoplasm.
4. There is much confusion concerning the ideas held by different investigators in regard to the nomenclature, the structure and the function of the various granules and crystals in the cytoplasm of the protozoa.
5. Biitschli's "Excretkornchen" consist of various different kinds of granules .and crystals found in protozoa. Schewiakoffs "Exkret- korner" are optically active bodies, nearly all of which, if not all, are crystals. Metcalf's "excretory granules" are not optically active. They are approximately IM in diameter and have staining properties like the mitochondria in other cells. The " Excretkornchen" of Biitschli, the " Exkretkorner' of Schewiakoff and the "excretory granules" of Metcalf therefore differ radically in structure.
6. Biitschli says the "Excretkornchen" are end products of metab- olism but he says nothing concerning their elimination. Schewiakoff contends that the "Exkretkorner" are formed in the food vacuoles, pass out into the cytoplasm, dissolve there and are eliminated by the contractile vacuole. Metcalf holds that the "excretory granules" ( Mast's beta granules) are permanent structures which function in excretion of substance by the contractile vacuole. Mast and Doyle maintain that they function in transporting substance through the cytoplasm.
LI'I I.K A I I KK CITED
BiiTSCHl.i, ()., 1881. Kxkrct knrnrr. Mronn's I >ii- Kl.i-^rn und Ordnungen des Tier-
reichs. Protozoa, lid. 1, S. KH. KHAINSKY, A., 1'Jll. Zur Morphologic und Physiologic emitter Infusorien (Para-
maecium caudatum) auf Grund einer neuen histologischen Methode. Arch.
f. Prolist., 21: 1-60. MAST, S. ()., 1026. Structure, movement, locomotion and stimulation in Amoeba.
Jour. Morph., 41: 347-425.
CONTRACTILE VACUOLE IN AMOEBA 1'koTKl'S (l.KIDY) 313
MAST, S. O., AND W. L. DOYLE, 1935a. Structure, origin and function of cytoplasmic
constituents in Amoeba proteus. I. Structure. Arch. f. Protist., 86: 155-
180. MAST, S. O., AND W. L. DOYLE, 19356. Structure, origin and function of cytoplasmic
constituents in Amoeba proteus with special reference to mitochondria and
Golgi substance. II. Origin and function based on experimental evidence;
effect of centrifuging on Amoeba proteus. Arch. f. Protist., 86: 278-300. MAUPAS, E., 1883. Contribution a 1'etude morphologique et anatomique des in-
fusoires cilies. Arch, de Zool. exper. et gen., Ser. 2, 1: 427-664. METCALF, M. M., 1910. Studies upon Amoeba. Jour. Exper. Zool., 9: 301-331. SCHEWIAKOFF, W., 1893. Uber die Natur der sogenannten Exkretkorner der Infu-
sorien. Zeitschr.f. wiss. Zool., 57: 32-56. TAYLOR, C. V., 1923. The contractile vacuole in Euplotes; an example of the sol-gel
reversibility of cytoplasm. Jour. Exper. Zool., 37: 259-290.
DURATION OF LIFE WITHOUT FOOD IN DROSOPHILA PSEUDOOBSCURA
OLE LILLELAND
(From the William G. Kerckhoff Laboratories of the Biological Sciences, ( 'lilifornia Institute of Technology, Pasadena, California)
Race A and Race B of Drosophila pseudoobscura have been dis- tinguished because they produce sterile male hydrids when crossed. Morphologically the races appear identical, and there are but few physiological characteristics that differentiate them. Thus, Poulson (1934) found that Race A flies have a shorter period of development than Race B flies. Dobzhansky l (1935) has studied the fecundity of both races at different temperatures and showed that during the lifetime Race A deposits more eggs than Race B at higher temperatures (25° C.), while at lower temperatures (19°, 14°) the relations are reversed. At all temperatures used Race B begins to oviposit later than Race A. In these experiments Dobzhansky observed also some indications suggesting that the longevity of Race A is greater than that of Race B. The present study is an attempt to secure further information bearing on this fact. In my experiments the measure- ments of the duration of life have been made in the absence of food. The data are therefore not necessarily comparable with those of Dobzhansky.
METHOD
I'earl and Parker (1924) state that any study involving longevity must be so conducted as to eliminate any desiccation; otherwise, the results are due to desiccation and not to the inherent factors which determine the length of life. Furthermore, they state that any technique which permits a fly to obtain water or any liquid even in small drops becomes a food study and not a starvation study. In the present work, houever. the interest centered in the reaction of the two races to varying humidii ies and temperatures and not purely in the length of life at one optimum condition. Undoubtedly at a humidity of 95 per cent the flies took moisture condensing on the side of the tube.
Five bottles each of Texas (A) and Seattle — 6 (B) cultures were
1 The writer wished to express his indebtedness and sincere appreciation to I )r. I )ol)/hansky.
314
DURATION OF LIFE WITHOUT FOOD IN DROSOPHILA 315
started on the twenty-ninth of May, each bottle containing 3 females and 5 males from stock that had been under laboratory conditions for some time. These were placed in the 19° (C.) cold room and trans- ferred every 48 hours to fresh bottles, until four successive transfers had been made. Any cultures which showed a total number of pupa1 under 100 or above 400 were discarded. On the fourteenth day of the culture's life, the paper was removed from the bottle and the pupae were scraped from it, using a pair of small forceps. Any pupae which were coated with food were discarded. The removed pupce were placed in one-ounce vials, plugged with cotton and placed again at 19° to await the first hatching.
Poulson (1934) found that the maximum hatching occurred between the hours of 6 and 10 A.M., excepting Race B females which emerged in greatest numbers between 2 and 6 A.M. Since a fair percentage of the latter also hatch between 6 and 10 A.M. these four
TABLE I
Duration of life (in hours) at densities of 5 and 30 flies per one-ounce vial. Tempera- ture 19° C., humidity 8 per cent.
|
Race |
Five per via |
1 |
Thirty per v |
ial |
Difference |
|
M ± m |
n |
M ± m |
n |
||
|
Seattle — 6 9 (B) |
71.70±1.18 |
00 |
69.60±1.10 |
60 |
+ 2.10±1.61 |
|
Seattle — 6<? (B) |
74.90±2.01 |
70 |
68. 80 ±1.44 |
60 |
+ 6.10±2.45 |
|
Texas 9 (A) |
102.57±1.20 |
105 |
97.0S±1.17 |
1.SO |
+5.52±1.67 |
|
Texas o" (A) |
104.88±1.76 |
7S |
98.67±1.33 |
90 |
+ 6.21±2.19 |
hours were chosen as the most convenient within which one might obtain representative populations. Each day of the hatch, therefore, all vials w7ere emptied of imagines at 6 A.M., and at 10 A.M. the flies which had emerged were etherized and placed in sterilized glass vials, 20 per vial, sexes segregated. Each vial was covered with a piece of medium muslin, held tight by a rubber band. The flies were allowed to recover from the effects of etherization before being placed in desiccators. Using Obermiller's (1924) table of vapor pressures over salt solutions, the following chemicals were used: concentrated H2SO4 for 0 per cent, KOH (fused) for 8 per cent, K2CO3-2H2O for 43 per cent, and K2SO4 for 95 per cent humidity.
After recovery the vials were numbered and placed upright in the desiccator, using a piece of white cardboard to hold the single row of vials close to the glass of the desiccator so that counts could be taken without removing the desiccator top.
OLE LILLELAND
RESULTS
An attempt was made to determine the length of life of D. pseudo- obscura at two population densities. Pearl and Parker (1924) found that, on food, densities from 35 to 45 flies per bottle were optimum for D. melanogaster . At lower and especially at higher densities a reduction of the life span was observed. The data for D. pseudoobscura are presented in Table I.
TABLE 1 1
Duration of life at two temperatures and four different humidities
|
Temperature, 24° Tcini><T.um.>. I1' |
Humidity 0 Per cent |
Humidity 8 Per cent |
Humidity 43 Per cent |
Humidity 95 Per cent |
||||
|
Race |
U ± m |
N |
M ± m |
.V |
M ± m |
N |
M ± w |
N |
|
tie — 6 9 ioia — 8 9 Texas 9 Mara — 3 9 |
70.38±1.91 63.10± .755 85.76±1.U> 67.45±1.14 |
60 60 60 129 |
77.64±1.62 65.12±1.77 97.20± .77 85. 80 ±1.93 |
120 161 55 130 |
94.28±1.01 si. 00 ±1.28 102.97 ± .(,<> 100.68±1.47 |
238 180 396 139 |
10 1.60 ±1.90 104.25 ±1.39 127.60± .'»s 110. 52 ±2.02 |
51 48 53 60 |
|
Seattle— 6 cf Sequoia — 8 cf Texas cf M.ir.i — 3cf |
74.0 ±1.39 66.75± .955 74.46±1.87 73. 57 ±1.02 |
60 60 60 110 |
74.40±1.14 66.37±1.1S <) 1.32 ±1.06 87. 00 ±1.88 |
131 159 100 97 |
99.10±2.415 s').22±1.20 104. 93 ± .80 99.84±2.10 |
210 160 254 120 |
92.40± .13 102.00± .86 111.0 ±1.33 113.10±1.90 |
70 40 40 40 |
|
-'it tie — 6 9 *->-<iuoia — 8 9 Texas 9 Mara — 3 9 |
^.16± .45 57.30± .945 35.58 ± .l')X |
100 80 100 |
53.52 ± .90 45.60± .733 63. 50 ±1.06 50.88± .657 |
120 80 100 220 |
55.16±1.80 57.08±1.37 73.84dbl.45 63 .60 ±1.63 |
100 80 60 80 |
(,(l •,(} • 1 HI 65. 10 ±.705 85 .80 ±1.67 74.40 ±2.36 |
40 100 40 80 |
|
1 1 le — 6 cf Sequoia — 8 cf Texas cf Mara— 3 cf |
37.80 ± .54 59.52 ±1.36 38.40 ± .39 |
60 60 80 |
50.79± .98 43.84± .857 62.57± .917 52.80± .622 |
120 80 100 200 |
54.68 ± .566 56.05 ±1.25 75.46 ±1.43 7 1.55 ±1.36 |
1 'ii 60 80 60 |
62.76± .115 71.66± .945 87.90 ±1.43 76.83 ±1.65 |
40 80 80 60 |
The data presented in Table I tend to show that the lower density is more favorable for D. pseudoobscura than the higher one. They show furthermore that the duration of life of Race A is greater than i hat of Race B. The latter conclusion must, however, be checked in several respects. First of all. the fact that one strain of Race A differs from one si rain of Race B does not necessarily indicate that the difference is characteristically racial; it may be a property of the par- ticular strains used. Moreover, a difference observed under one set of conditions may be obliterated or even reversed under other con- ditions. The longevity of the flies has therefore been studied at temperatures 19° and 24° C., and at humidities of 0 per cent, 8 per cent, 43 per cent and 95 per cent (Table II). Finally, a variety of strains of both races coming from different parts of their distribution area has been studied (Table III).
Among the strains used, Texas, Mara, Julian, Oaxaca, Olympic, Pavilion, and Taos belong to Race A, and Seattle, Sequoia, Campbell, Crater, Ouilcene, and Ouesnel to Race B. It can be seen from Tables
DURATION OF LIFE WITHOUT FOOD IN DROSOPHILA 317
II and III that in general Race A tends to be more long-lived than Race B. Nevertheless, some intra-racial variation is observed. Thus, the longevity of the Texas strain is greater than that of the Mara strain; the Quesnel strain is superior to other Race B and to some Race A strains. It may be noted that at 95 per cent and 43 per cent humidity Race A strains are in general superior in longevity to Race B. At lower humidities, especially at 0 per cent, a greater degree of overlapping between races is observed. Qualitatively, the
TABLE III
Duration of life in nine strains of D. pseudoobscura at 24° C. and three different
humidities
|
Humidity 0 Per cent |
Humidity 43 Per cent |
Humidity 95 Percent |
|||||
|
Race |
M ± m |
* |
M ± m |
N |
M ± m |
N |
|
|
Campbell— 4 9 |
40.13± .770 |
90 |
48.85±.615 |
80 |
68.57 ±1.40 |
80 |
|
|
Crater— 2 9 |
38.09± .628 |
60 |
46.80±.514 |
60 |
61.86±1.05 |
60 |
|
|
Quilcene — 4 9 |
40.00 ± .740 |
90 |
47.04±.648 |
60 |
66.24±1.68 |
40 |
|
|
*2 |
Quesnell— 5 9 |
46.91 ± .655 |
60 |
56.12±.727 |
68 |
69.43±1.585 |
60 |
|
oj |
Campbell — 4 9 |
41.02± .634 |
103 |
51.64±.605 |
80 |
66.62±1.54 |
60 |
|
Crater — 2 cf |
38.10± .715 |
60 |
50.00±.595 |
60 |
62.15±1.095 |
82 |
|
|
Quilcene — 4cf |
38.84± .500 |
120 |
43.08±.468 |
60 |
57.30±1.49 |
60 |
|
|
Quesnell — 5 of |
47.07 ± .475 |
60 |
55.41 ±.495 |
80 |
59.10±1.295 |
60 |
|
|
Julian E— 6 9 |
53.02± .770 |
80 |
63. 72 ±.755 |
60 |
83.62±1.96 |
60 |
|
|
Oaxaca — 4 9 |
43.08± .707 |
85 |
50.64±.741 |
60 |
90.54±1.86 |
80 |
|
|
Olympic — 2 9 |
46.62 ± .684 |
80 |
61.58±.787 |
40 |
92.05±1.69 |
80 |
|
|
Pavillion— 5 9 |
49.25 ± .705 |
89 |
66.37±.455 |
100 |
88.88±1.54 |
60 |
|
|
^ |
Taos— 1 9 |
49.89±1.18 |
70 |
56.40±.800 |
60 |
70.10±1.54 |
80 |
|
OJ |
|||||||
|
u |
|||||||
|
02 |
Julian E — 6cf |
47.02± .655 |
100 |
60.00±.482 |
80 |
76.44±2.21 |
60 |
|
Oaxaca — 4 cf |
45.32 ± .644 |
100 |
56.57±.616 |
60 |
90.73±2.41 |
60 |
|
|
Olympic — 2cf |
46.85 ± .632 |
90 |
65.43±.880 |
40 |
92.93±1.68 |
80 |
|
|
Pavillion — 5 a* |
50.00± .806 |
60 |
64.87 ±.545 |
80 |
89.73±2.01 |
40 |
|
|
Taos — 1 cf1 |
51.78± .955 |
60 |
61.60±.641 |
80 |
64.56±2.00 |
60 |
effects of humidity and temperature are identical for all strains studied. The duration of life is greater at the lower temperature (19°) than at the higher one (24°), and at higher humidities than at the lower ones. The two races of D. pseudoobscura differ in their geographical distribution. Race B is restricted to the northern part of the Pacific Coast, while Race A lives much further eastward and southward than Race B (Dobzhansky, 1935, 1937). As far as our data show, the effects of humidity and temperature on the duration of life in the absence of food have little bearing on the geographical distribution of
318 OLE LILLELAND
the two races. Further studies, especially those on the duration of life in the presence of food, may conceivably throw more light on this problem.
SUMMARY
1 . In the absence of food, the duration of life of Drosophila pseudo- obscura is greater at lower temperatures, greater humidities, and lower population densities studied.
2. With temperature, humidity, and population density being kept constant, Race A lives longer than Race B. These differences between the races is more pronounced at higher than at lower humidities.
LITERATURE CITED
DOBZHANSKY, TH., AND R. D. BocHE, 1933. Intersterile races of Drosophila
pseudoobscura Frol. Biol. Zentrlbl., 53: 315. DOBZHANSKY, TH., 1935. Fecundity in Drosophila pseudoobscura at different
temperatures. Jour. Exper. Zoo/., 71: 449.
DOBZHANSKY, TH., 1937. Genetic nature of species differences. Am. Nat., 71: 404. OBERMILLER, J., 1924. Equilibrium humidities over saturated salt solutions.
Zeitschr.f. Physik. Chem., 109: 145. PEARL, R., AND S. L. PARKER, 1924. Experimental studies on the duration of life.
X. The duration of life of Drosophila melanogaster in the complete absence
of food. Am. Nat., 58: 193. POULSON, D. F., 1934. Times of development of the two races of Drosophila
pseudoobscura. Jour. Exper. ZooL, 68: 237.
THE LIFE HISTORY OF A TREMATODE (LEVINSENIELLA CRUZI?) FROM THE SHORE BIRDS (LIMOSA FEDOA AND CATOPTROPHORUS SEMI- PALM ATUS INORNATUS) l
(Contributions from the Scripps Institution of Oceanography New Series, No. 19).
R. T. YOUNG (From the Scripps Institution of Oceanography, La Jolla, California)
INTRODUCTION
Lebour (1911) has described a cercaria (C. ubiquita] from British snails (Palndestrina stagnates, Littorina obtusata and L. rudis). She found "Cercarise exactly corresponding with C. ubiquita, but without the glands, inside the tissues of Carcinas mcenas" and similar meta- cercarise were found by her in Cancer pagurus, the third host in the cycle being presumably the herring gull (Lams argentatus}. The same or a closely related larva has been found by Stunkard (1932) in Littorina rudis and L. littorea at Roscoff, France, and Rees (1936) has observed similar larvae in these snails and in L. obtusata at Aberystwith, England. The adult form of these larvae is supposed to be some species of Levinseniella, but thus far the various steps in the life history have not been definitely established.
In a study of the trematodes of the La Jolla region on which I am engaged I have found a stylet cercaria infesting the snail Olivella biplicata which differs in some particulars from those of Lebour and Stunkard, and whose probable adult form I have located in the shore birds (Limosa fedoa and Catoptrophorus semipalmatus inornatus], while the metacercaria parasitizes the sand crab (Emerita analoga).
I will first describe the cercaria, the metacercaria and the adult worm, and then give the evidence on which my conclusions regarding the life history are based.
THE CERCARIA
The cercaria (Fig. 1) is one of the spineless xiphidio-cercarise, possessing a stylet from 24 to 32 ju long, and a set of eight unicellular penetration (?) glands, but no pigment spots. In fixed specimens the body averages 165 X 46 n and the tail 124 /* in length. It has an oral
1 The following study has been conducted at the Scripps Institution of Ocean- ography, to whose directors, Dr. T. Wayland Vaughan and Dr. H. U. Sverdrup, and other members of the staff I am indebted for many courtesies.
319
320
R. T. YOUNG
sucker 34 n in average diameter, but no acetabulum, or gut. The excretory bladder is bifurcate, but further details of the excretory system are not evident.
Perhaps the most interesting feature of the larva are the penetra- tion (?) glands, which are in two sets, an anterior set of four larger, and a posterior one of four smaller glands. Their arrangement is shown in Fig. 1, where they are designated by numbers 1-8 respec- tively. Kuch gland has a duct which follows a more or less sinuous course depending on the state of contraction of the worm, and which opens a i ihe anterior border of the >ucker. Lebour states that then-
I i'.. 1. Cercaria showing stylet, sucker, nerves, excretory bladder and glands (1-S). I or dea lipii-m of latter, see u-\t. Outline with camera, details free-hand.
are but four ducts for the group of -lands in her larva, bill their close apposition and sinuous course renders their individual identification difficult and mighi easily lead to confusion about their actual number. My specimen^, \\heu stained with neutral red, show clearly one duct for each gland. lioih glands and ducts Main intensely in this stain.
These glands differ markedly; not only in sixe, but also in the physical and apparently chemi. a! character of their contents. Glands No. 1 and 2 are finely granular and do not ordinarily take an eosin stain. Occasionally Nos. 1 and 1 slain lightly in eosin but much less
LIFE HISTORY OF A TKKMATODK
321
intensely than Nos. 3 and 4. Numbers 3 and 4 are also finch' granular, though the granules are perhaps a trifle coarser than in Nos. 1 and 2, but they take the eosin stain intensely. Glands 5-8 stain very lightly. They contain only a few eosinophile granules which are scattered through the meshes of a rather open reticulum, so that the whole cell appears as an open space in the sections with what is usually a pycnot ic.
FIG. 2. Photomicrograph of long section of cercaria showing glands 2, 4, 6 and 8 (see Fig. 1 and text), d, ducts; e, excretory bladder; X 860.
heavily stained nucleus near the center. The nuclei in all eight glands, however, vary widely, sometimes being pycnotic and densely stained, and at other times vesicular and so lightly stained as to be scarcely visible. Especially is the latter true of glands 1 and 2. That these differences are real and not artificial is indicated by the fact that they may occur in neighboring larvae in the same sporocyst. These struc- tural and staining differences are evidence of functional differences in the two sets of cells 1-4 and 5-8 respectively. It is probable that each set elaborates a different secretion, for were the differences in structure
322 R. T. YOUNG
and staining capacity due to different stages in the elaboration of the same secretion the two sets of cells should vary in different individuals. But this is apparently not the case; each set is distinct in every cercaria. The difference in eosinophily between cells 1, 2, and 3, 4 respectively is probably explicable by differences in the stage of development of the same secretion, but this will not explain the dif- ference between sets 1 4 and 5-8.
Stunkard (he. cit., p. 337) says of his larva that "the cells do not stain with neutral red, but the secretory granules are clearly visible . . . (\\hich granules) stain a deep blood red in the terminal portion of the ducts and frequently accumulate there to form enlargements," while Rees (loc. cit., p. 622) says that "only the two anterior pairs (of cells) and their ducts . . . (stain) a deep red. The remaining cells and their ducts have never been observed to take up the intravital stain." All eight cells in my larva stain uniformly in neutral red, the differences appearing in the material stained with eosin.
It is difficult to harmonize these various observations with one another. They may indicate specific differences in the cercariae them- selves but it is not impossible that they are due to functional differences in the glands at different times and under different conditions. Possibly they are the effect of different hosts.
The infestation of the snail is considerable, 6 out of 98 specimens examined being infested.
The shedding of the cercariae appears to be a seasonal matter, although other, as yet obscure factors may affect this. In February and March, 1935, cercariae were easily obtained by isolating snails from the beach near the Scripps Institution, whereas from November, 1936 to January, 1937 isolation of several hundred snails produced no cercariae. In May and June of 1937, a few cercariae were obtained from several hundred snails, but in July and August, 1937, isolation of -everal hundred snails failed to produce any. That some of the apparently sterile snails \\ere infested, however, was proven by dis- section of several specimens. Lebour (loc. cit.) also has observed that .-nail> are free from ('. i(hi(/nitn in winter, while their walled cysts (metacercarise) are present in crabs only during spring and summer and early aiiinmn. I have not made sufficient study of this matter, however, and hence cannot define definitely the seasonal cycle of the life history, it Midi exists, which appears to be the case.
Till', Ml I Ml KCAKIA
The cercai i. i- in the crab lack tails, but \\ hether these are lost before or after entrance I cannot say. The foinier appears more likely,
I.II-K HISTORY Oh' A TKEMATODE
liowever, as the tails are frequently detached from the cercariae while swimming. Soon after entrance the glands and ducts become broken down and are identifiable only as irregular masses which still stain in neutral red. Within two or three days the cercaria secretes a cyst within which it curls up and begins its development into the meta- cercaria.
This naturally varies in size with age, the older ones reaching a diameter of 0.3 mm. with a cyst wall about .025 mm. thick. It is very slightly ovoid in shape and is found in the connective tissue of the crab, chiefly among the liver tubules. Practically every crab examined, in the La Jolla region, over 6 mm. in length is infested. I have not determined the maximum number of cysts which one crab may carry, but it is probably several dozen.
THE ADULT
The adult worm agrees closely in all important details with the description of the genus Levinseniella as given by Jagerskiold (1900). The outline is narrower in front and broader behind, and when much expanded the body may be slightly constricted just in front of the acetabulum. This latter is located about two-thirds the length of the body from the anterior end. The anterior two-thirds of the body is spinose, the spines gradually diminishing in size and disappearing at about the level of the acetabulum.
Judging from the figure (unlabeled) and the description (in Portu- guese) of L. cruzi by Travassos (1920) my specimens belong to this species, although the difference in the orders of the hosts (Anseres and Charadriformes) and in locality (California and Rio de Janeiro, Brazil) render this uncertain. As shown in Table I, the dimensions of my specimens correspond rather well with those of L. cruzi,- the chief differences being in the size of the oral sucker, which in my specimens averages 0.061 mm. in diameter while in the former it averages from 0.092 to 0.113 mm., and in the length of the ceca (0.134 and 0.163 to 0.241, respectively).
Perhaps the most striking feature of agreement between the two is in the size of the eggs, which average larger in both than in any other species known to me. The average of 24 eggs from my speci- mens fixed in aceto carmine is 0.036 by 0.017 mm., while in L. cruzi they average from 0.021 to 0.035 by 0.012 to 0.014. When fairly extended in life my specimens range from 0.55 to 0.9 mm. in length, while L. cruzi ranges from 0.609 to 0.713 mm. While there thus appear to be certain minor differences between my worms and those of
2 Presumably measured from fixed material although Travassos does not so state.
324
R. T. YOUNG
TAHLK I
Table I showing; average measurements in M of fresh and preserved specimens, with the number of specimens averaged for each measurement in parentheses. With the exception of five eggs, all preserved specimens were fixed in aceto-carmine. There is some overlap in the two sets, i.e. a few specimens were measured first in the fresh, and then in the prc-crvcd condition. Column 3 shows the measurements of L. cruzl from Travassos (loc. cit.).
Fresh
Preserved
L. cruzi
Length.... 692
(14)
Width. 193
(14)
( ir.il -ucker 61
(12)
Pre-pharyn.\ 37
(3)
Pharynx 33X21
(8)
Post-pharynx 168
(4)
Ceca. 112
(7)
Acetabulum 54
(15;
.V ctab. distance (from ant. end). 485
(9)
is diameter 23
(9)
Scinin.il \oiclc 67X39
(5)
Ovarj 74X49
(12)
Kiuht testis 97X58
(12)
I. it 90X61
(12)
676
15
303 (15)
61
(14)
609-713
222-313
92-113
36
(5)
32X23
(9)
170
(10)
134 (10)
55
(15)
457 (9)
23
(9)
67X38 (8)
71X54
I5i
14-56
35X53-21X35
63-120
163-241
56-71
56-92
93X69
(14)
89X65
(14)
36X17
(24)
35X14-21X12
LIFE HISTORY OF A TREMATODE
Travassos, they are insufficient in my judgment for separating them specifically, and I am therefore, in spite of the difference in hosts and locality, provisionally assigning my specimens to his species.
THE LIFE HISTORY
The evidence upon which my conclusions regarding the life history of this worm are based is briefly as follows:
(1) the presence in the sand crabs of stylet cercariae, and of all stages in the developing metacercariae from that of the cercaria sur- rounded by a very thin cyst, but still having the stylet and remnants of the penetration (?) glands to those of the trematode, which resembles the adult in every way save for the presence of eggs in the uterus;
(2) the experimental infestation of crabs with cercariae;
(3) the presence in the birds of adult worms which resemble in every way the excysting metacercariae, save for the presence of eggs in the uterus, and the presence of both encysted and excysted meta- cercariae in the birds' stomachs after they have fed on sand crabs.
There is apparently in nature a universal infestation of all crabs more than 6 mm. long. I have not recorded the number of crabs which I have examined (50-100), all of which were infested. It is probable that infestation begins when the crabs are about 5 mm. long as I have found specimens of this size containing cercariae. Until they reach a length of at least 5 mm. the crabs are less frequently infested. Of 30 specimens of this size or less which I examined in nature only 8 contained cercariae. It is evidently a matter of age, and consequently length of time of exposure, which determines the infestation, for the larger specimens are more heavily infested than the smaller ones, and very young crabs are susceptible to experimental infestation. The season of the year is possibly also an important factor, for, as already shown,3 the deposition of cercariae by the snails may be dependent on the season.
I have performed three sets of experiments in an attempt to demon- strate the transfer of cercariae from snail to crab, the results of which are summarized in Table II.
In the first of these a number of cercariae were obtained from snails which shed them naturally, transferred to a finger bowl with ten crabs from 4 to 5 mm. long, and left for from three to four days, after which, examination of the latter showed that seven out of eight (all of which were examined at the time) contained from one to four cercariae each while the remaining crab which was dissected 42 days later was negative.
3 See page 322.
326 R. T. YOUNG
As a control 15 crabs collected on the beach within four weeks of this experiment gave a ratio of 5 infested to 10 non-infested specimens.
An experimental infestation of seven out of nine young crabs, while 10 of 15 in nature were uninfested, has a probability, on the basis of chance alone, of . 03333. 4
A second group of experiments consisted in holding for one or two days 18 young I'.nierita in finger bowls, together with cercarise which had just been shed by snails. Seven of these were subsequently found to be infested. In the control for this experiment, out of 20 young crabs none were infested. In this experiment and its control the crabs were kept for several days previously in running sea water which was filtered through several inches of sand, thereby eliminating the likeli- hood of an accidental infestation. The probability of this result occurring on the basis of chance alone is .002 1.4 In explanation of
T AIU.K II
Showing results of experimental infestation of Emcritu with the cercaria of
Levinseniella cruzi (?) 6
Series 1 Series 2 Series 3
Experiment Experiment Experiment
Emerita exposed to l-.nii-ritn exposed to Emerita exposed to
cercariae cercari.r Olirrlld
Positive Negative Positive Negative I'o-itive Negative
72 7 11 17 18
Control Control Control
Emerita collected on l-'.nu'rita in filtered Emerila in un filtered
beach sea water sea water
Posit ive Negative Positive Negative Positive Negative
5 10 0 20 6 40
the low ratio of infestation (7 : 18) in this series of experiments, as < -'.mpared with the high ratio (7 : 9) in the former series, it may be sii<l that the number of cercaria per crab was much smaller in the l.ittrr than in the former series.
In a third scries of experiments Emerita were held from one to several days in jars6 containing infested snails for comparison with crabs held in running sea water. The result of this experiment was an infestat ion of 1 7 out of 35 crabs in the experimental jars compared with 6 out of 46 in the control. The probability of this result on a purely chance basi^ is .0003.7
4 Computed from M< K\u-n (1929).
k In determining the pri-srm i- or absence of parasites in the crabs the factor of length of examination is rssrnn.il. I adopted a minimum of 12-15 minutes for the smaller (5-6 mm.) specimens and of 15-20 for the larger (10-15 mm.) ones.
6 Some of these were supplied with running sea water. In others the water was standing, but aerated.
7 Computed from McEwen (1929).
LII-K HISTORY ()!• A TKKMATODE
In evaluating this series of experiments it should be observed that in many of them only a very meagre number of cercariae were furnished a considerable number of crabs in a volume of from two to eight liters of water, hence the chance of any individual crab receiving a cercaria was rather small. The infestation of a considerable number of crabs in the controls (6 in 46) is difficult to explain. Very rarely cercariae occurred in the laboratory water supply, but straining through cotton a small stream of water from a tap for about forty hours pro- duced no cercaricc, so that their chance occurrence in the control aquaria can hardly be explained thus. In many cases the control aquaria were standing on the same table with jars which contained possibly infested snails and it is possible that cercaria? from the latter may have infested the controls. In one case in particular a control jar of Emerita collected on August 17, 1937 contained recently infested crabs until August 28, during which time a jar of snails was standing on the same table; but straining the water from this jar through fine bolting cloth on this date failed to produce any cercariae. On August 30 and September 1 after removal of this jar an examination of ten crabs from this control revealed no recent infestations.
The source of the occasional infestation of the controls therefore remains uncertain, but in any case the difference in the infestation ratios between experiments and controls in this series is highly sig- nificant.
Taken collectively the experiments give a result of 31 infestations out of 62 tests, while the controls give 11 out of 81 tests; and these include 5 out of 15 exposed to natural infestation on the beach. This result, on the basis of chance alone, has a probability of .0000013.8
Larvae similar to those in the sand crabs occur in the fiddler crabs (Uca crenulata) from Mission Bay, an inlet of the sea near La Jolla. Doubtless, many other crabs of the vicinity harbor the same or related larvae, for Lebour (loc. cit.) and Stunkard (loc. cit.) in Europe have found related metacercariae in several species of crabs.
The extent of infestation of the willet and godwit by this parasite is rather limited. Only toward the end of this study did I differentiate between this species and two other related ones so that I cannot give any figures on the percentage of infestation. However, the former is usually, if not always, greatly outnumbered by what appear to be two other species of Levinseniella. In at least one case, that of a marbled godwit, it was apparently absent, but time did not permit an examina- tion of the entire gut of this bird so that possibly a few specimens escaped observation.
8 Computed from McEwen (1929).
328 K. T. YOUNG
McMulIcMi (1935) gives some data which are of interest in this con- nection. In experimental infestations of Amia calva with Macro- deroides typicus he found the fish heavily infested a few days after feeding infested material, but after several months the infestation was only slight. Thus one fish examined after 22 days contained by actual count at least IS'U flukes, while another fish examined 202 days after infestation contained only 33. He believes that 'This enormous loss . . . was probably . . . due to crowding" (he. cit., p. 375). And Sarles and Stoll (1935, p. 290) state that "cats . . . carrying natural infestations of Toxacara cati were found to possess a uniformly high degree of resistance against attempts to superimpose infection with this ascarid." What limits the degree of infestation in wild animals, many of whom, as is the case with these shore birds, have unlimited opportunity for acquiring it, is as yet an unsolved problem. It may be, as suggested by Stunkard (1930) and Stoll (1929), analogous to the development of immunity by animals to bacterial infection. In tin- case of L. cruzi and the shore birds it is possible that the usually abundant infestations with other species limits the number of the former. It does not, however, apparently interfere with extensive infestation by a Maritrema sp. which the birds probably receive from the sand flea (Orchestoides) , which forms a considerable part of their diet.
In trying to follow the life history of this worm I was at first misled by finding excysted individuals in the surf perches (Embiotoca, Abeona) after the latter had eaten sand crabs. It was only after performing several experiments that I discovered that the infestation of the fish is only temporary. It is rather remarkable, however, that a parasite of a homoiothermal animal should excyst and exist, even lor a brief period in a poikilothermal form. My observations on this point are not solitary, however, for Linton (1928) reports Levinseniella ad nn«i from a sanderling (Crocethia alba} which "appeared to be identical with a species found in the load fish (Opsanus tail)" (loc. <it., p. 21).
The last link in the life history of this worm is not yet made. In spite ol n-pi-aied fecal examinations of both wild and captive birds, I have not nun-eded in finding either eggs or miracidia. The eggs in utero appear- never to have de\rloped beyond an early cleavage stage. I have one probable record of a miracidium in the snail, but as my observations were very brief, the organism being lost in transfer to a slide, I am not able to assert positively that it was a miracidium, much less to determine whence it came. There must, however, be
I. IFF. HISTORY OF A TRKMATODK
some period when the birds are shedding eggs or miracidia. This I hope subsequent study will reveal.
SUMMARY
The life history of a trematode (Levinseniella cruzi?) from the shore birds (Limosa and Catoptrophorus] and the structure of its cercaria have been described in some detail. The probable life history is as follows :
The miracidium has not yet been certainly observed. The cercaria, which belongs to the ubiquita group of Lebour, inhabits the snail, Olivella biplicata. From here it passes to the sand crab, Emerita analoga, which latter, when eaten by the birds, infests them with the eidult trematode. The degree of infestation of the snail is considerable, but the cercariae are shed infrequently, possibly at definite seasons.
Practically all crabs over 6 to 7 mm. in length are infested, but for some as yet undetermined reason, the infestation of the birds is light.
BIBLIOGRAPHY
JAGERSKIOLD, L. A., 1900. Levinsenia (Distomum) pygmeea Levinsen, ein genital-
napftragendes Distomum. Centralbl.f. BakL, Aht. 1, 27: 732. JAGERSKIOLD, L. A., 1909. Kleine Beitrage zur Kenntniss der Vogeltrematoden.
Centralbl.f. BakL, Abt. 1, 48: 302.
LEBOUR, M. V., 1911. A review of the British marine cercarise. Parasitol., 4: 416. LINTON, E., 1928. Notes on trematode parasites of birds. Proc. U. S. Nat. Mus.,
73: 1. McEwEN, G. F., 1929. Methods of estimating the significance of differences in,
or probabilities of fluctuations due to random sampling. Bull. Scripps Inst.
Oceanography, Tech. Ser., 2:1. McMuLLEN, D. B., 1935. The life histories and classification of two allocreadiid-like
plagiorchids from fish, Macroderoides typicus (Winfield) and Alloglossidium
corti (Lament). Jour. Parasitol., 21: 369. REES, W. J., 1936. Note on the ubiquitous cercaria from Littorina rudis, L. obtusata
and L. littorea. Jour. Mar. Biol. Ass'n., 20: 621. SARLES, M. P., AND N. R. STOLL, 1935. On the resistance of the cat to superimposed
infection with the ascarid, Toxocara cati. Jour. Parasitol., 21: 277. STOLL, N. R., 1929. Studies with the strongyloid nematode, Haemonchus contortus.
Acquired resistance of hosts under natural reinfection conditions out-of- doors. Am. Jour. Hygiene, 10: 384. STUNKARD, H. W., 1930. The life history of Cryptocotyle lingua (Creplin), with
notes on the physiology of the metacercariae. Jour. Morph., 50: 143. STUNKARD, H. W., 1932. Some larval trematodes from the coast in the region of
Roscoff, Finistere. Parasitol., 24: 321. TRAVASSOS, L., 1921. Contribuic.ao para o conhecimento da fauna helmintologica
brazileira IX, Sobre as especias da subfamilia Microfalinae Ward 1901.
Arch. Escol. Sup. Agr. Med. Vet. Nictheroy, 4 (1920) 1921, pp. 85-91.
THE EFFECTS OF ILTRA-YIOLET RADIATION OF X2537A I POX CLEAVAGE OF SEA URCHIN EGGS1
ARTHUR C. GIESE
ruin the School of Biological Sciences, Stanford University)
The effect of ultra-violet radiation upon the division rate of various cells has been the subject of a number of investigations, but the results are not in complete agreement, some investigators claiming acceleration (e.g., Coblentz and Fulton, 1924; Hinrichs, 1928; Alpatov and Nastjukova, 1933), others retardation (e.g., Hertel. 1905; Gates, 1()29; Oster, 1934; Chase, 1937), still others acceleration or retardation depending upon the \\ave-length and dosage (e.g., Bovie and Hughes, 1918; Hughes and Bovie, 1918; Hutchinson and Ash ton, 1929).
The disagreement may be more apparent than real, the results depending upon the wave-length of the ultra-violet, the dosage and the organism used. Quantitative data are needed to throw further light upon the problem. The following paper is an attempt to gather such data on the effects of one wave-length, 2537A, upon cleavage of sea urchin eggs, this material being chosen because the self-contained food supply greatly simplifies control of the environment. \York is planned at each of the other wave-lengths of ultra-violet light represented in the spectrum of the quartz mercury arc.
MATERIALS AND METHODS
The sea urchins (Strongylocentrotus pnrpiimtns Stimpson) were collected during the winter breeding seasons of 1935 and 1936 at Moss Beach and Pacific Grove, California. Eggs were obtained after natural spanning or by excision of the ovaries and from 25 to 100
s \\ere placed in each 1" watchglass containing 3/4 cc. sea water. The ^penn suspension was determined each time by tests, successive dilution^ being made until 100 per cent fertilization was achieved without overinsemination, and since almost all eggs cleaved normally, polyspermy was probably rare. Only eggs in which practically 100 per cent showed fertilization membranes within two minutes after insemination were used in this research.
In the case of eggs kept at 14 16"" ('., the first cleavage occurred in about two hours following insemination, the second after another
1 This work lias been supported in part by funds granted by the Rockefeller Foundation.
ULTKA-YIOLKT RADIATION AND CLEAVAGE
hour, and the successive cleavages to the 64-celled stage at about hourly intervals. Within 20 hours after insemination the actively swimming blastulac had been formed and in about another 24 hours gastrulation was complete, having begun some 12 hours earlier. Plutei formed within 4 days after insemination although the arms did not begin to elongate for another 24 hours. Since controls showed normal development "to the pluteus, the conditions were considered satisfactory for the investigation.
For the first series of studies a mercury-argon discharge tube which emits about 88 per cent of the total output in the visible and ultra- violet at X2537A (Coblentz, 1931; Leighton and Leighton, 1935) was employed since its high intensity enables one to give an effective dose of radiation in a short time and renders feasible certain experiments otherwise impossible. The intensity of the radiations, after screening out the infra red rays by a suitable water filter, was showrn by thermo- pile measurements to be relatively constant for the period of investi- gations.
For experiments where pure light of X2537A was needed the radiations from a water-cooled quartz mercury arc were passed through a natural quartz monochromator and the light of the desired wave- length was focused on the quartz cell containing the eggs. The apparatus used was in general similar to that previously described (Giese and Leighton, 1935).
The line thermopile (type described by Leighton and Leighton, 1932, p. 1884) used in series with a D'Arsonval H.S. galvanometer, was calibrated against Bureau of Standards Lamps C-211 and C-212. The thermopile factor for X2537A was calculated to be 24.10 ergs /sec./ cm. galvanometric deflection.
EXPERIMENTAL Irradiation of Eggs Just Before the First Cleavage
In the first series of experiments the eggs which had been insemi- nated 90 minutes previously were irradiated for 1, 2, 4, 8, 16, 32, and 64 seconds at a distance of 31 mm. from the center of the mercury argon tube, and for 1, 4 and 16 seconds at a distance of 248 mm. from the center of the tube (to give exposures approximately equivalent to 1/64, 1/16, 1/4 second). In one series exposures approximately equiva- lent to 1/8 and 1/32 second were also given. Examinations were made at intervals of a half hour or an hour, depending upon circumstances, and the stages in development recorded. A typical set of data from a series of 3 experiments is plotted in Fig. 1.
332
ARTHUR C. GIESE
It is clear from Fig. 1 (hat delay in cleavage is proportional to dosage and that even 1 second irradiation is sufficient to delay develop- ment for a long time, 1/4 second for a slight period of time. However, the protoplasm of the egg hastily repairs injury from lesser dosages, for example, the rate of cleavage of eggs irradiated 1/8, 1/16, 1/32 and 1/64 second was at no time lower than that of controls. But in no case was an incn MM <1 rate of cleavage observed.
1
I i
123456 7 8 9 10 II
TIME SINCE INSEMINATION (HOURS)
FIG. 1. Retardation of cleavage by radiation.
A. Control.
/>'. One-fourth of a second dosage.
C. One second dosage.
/'. Two seconds dosage.
/'.'. Four seconds dosage.
I'". Eight seconds dosage.
G. Sixteen seconds dosage.
//. Thirty-two seconds dosage.
\\ hilc main- of tin- eggs in .idiatcd for 4 or more seconds developed normally later, in most cases a retardation was observable even at later stages in development. Thus in Table I it will be noted thai while 22 hours after insemination normal free-swimming blastuhe were formed from all eggs except those irradiated 4 seconds or more, those developed from eggs irradiated 4 and 8 seconds were still within
ULTRA-VIOLET RADIATION AND CLEAVAGE
333
the fertilization membranes. Even 32 hours after insemination gastrulation in these eggs was just beginning when the controls and those given smaller doses had completed invagination. Ultimately all the above -eggs, even those irradiated as long as 8 seconds, gave rise to gastruloe normal to all appearances.
TABLE I
Later development of eggs irradiated ninety minutes after insemination at the
dosages indicated
|
Dosage in seconds |
22 hours after insemination |
32 hours after insemination |
46 hours after insemination |
|
1/64, 1/32, 1/16, 1/8, 1/4 and con- trol |
Normal free-swim- ming blastulae |
Normal gastrulse |
Beginning of gut differentiation |
|
1 |
Normal free-swim- ming blastulae |
Normal gastrulae |
Gut not yet dif- ferentiated |
|
2 |
Normal free-swim- ming blastulae |
Normal gastrulae |
Gut not yet dif- ferentiated |
|
4 |
Normal blastulae but still within mem- branes |
Blastulae with begin- ning of invagination |
Early gastrulte |
|
8 |
Normal blastulae but still within mem- branes |
Blastulae with begin- ning of invagination |
Early gastrulae |
|
16 |
Mostly abnormal mo- tile balls. Some non- motile blastulae. |
Mostly abnormal. Show delayed gastru- lation. |
As before |
|
32 |
Coagulated cells |
— |
— |
|
64 |
Coagulated cells |
• — |
— |
In some series (5 trials) of eggs irradiated 16 seconds a fair propor- tion of the eggs formed blastula? and gastrulae; in other series of eggs so irradiated cleavage resulted in a mass of motile cells which persisted without differentiation for as long as observations were made. After dosages of 32 seconds eggs developed into masses of undifferentiated cells which never became motile and early appeared opaque. Dosages of 64 seconds apparently inhibited even one division, though in some cases cleavage did occur; more often the attempt at cleavage was abortive, the apparent blastomeres failed to separate, often fused and then cytolized.
334
ARTHUR C. GIESE
Differential Susceptibility of Eggs at Different Stages
It would be of interest to compare the susceptibility of unfertilized eggs, eggs just inseminated, eggs well after insemination, and eggs in the first cleavage. First of all, however, it was necessary to determine whether the unfertilized egg would become activated with the dosages used. Three series of 3 trials each were therefore made with 1/64, 1/16, 1/4, 1, 16, 64, and 256 seconds of irradiation. In no case was a
Control. Unfertilised i'lfteen iLlriutf* after Ir.-.erolnntion irinutes r.ftcr inseriii;tt Ion. Two-celled str»ge.
45678 TIME S/NCE INSEMINATION (HOURS)
(0
I ic.. 2. Differential susceptibility to radiation of eggs at different stages of development .
A. Control.
B. One second dosage.
C. Sixteen seconds dosage.
normal fertilization membrane formed without insemination, nor was division apparent. The dosages used do not, then, activate the eggs to artificial parthenogenesis, but experiments showed that when the dosage was not too great, eggs so irradiated when inseminated would develop normally.
Three series of experiments (3 each) were now performed on eggs in the following stages: (1) unlCi I ili/cd, (2) 15 minutes after insemina-
ULTRA-VIOLET RADIATION AND CLEAVAGE
tion, (3) 90 minutes after insemination and (4) the two-celled stage with dosages of 1, 4, 16, and 64 seconds of irradiation. The results, excepting those for eggs irradiated 4 and 64 seconds, are given in Fig. 2. It is readily observed that there is no very great difference of sus- ceptibility of these stages, one second of irradiation reducing the cleavage rate about equally in all cases (4 seconds reducing the rate still more, but omitted from the graph to avoid confusion), 16 seconds much more. From Fig. 2 it may be observed that there is a latent period before the effects appear. This is particularly so for all stages irradiated for one second and for the 2-celled stage irradiated at all dosages. The eggs irradiated 64 seconds gave but few abortive cleavages when irradiated before insemination or 15 minutes after insemination. When irradiated for 64 seconds, 90 minutes after insemination, most of the eggs passed into the 2-celled stage, and then cleaved abortively. Apparently at the time of irradiation the mecha- nism of cell division was already in full swing and could not be stopped. Of eggs irradiated for 64 seconds just after the first cleavage only a small proportion continued to divide and these only for a short time afterwards.
As in the previous experiments retardation was not only obvious during early cleavage but also at later stages, for when controls had developed into free-swimming blastulse, some of the irradiated eggs had developed into non-motile blastulae, and when controls had gastrulated, some of those irradiated had only begun to gastrulate. The data for later development are given in full in Table II.
Three sets of eggs in the 2-celled stages were irradiated with dosages equivalent to 1/4, 1/16, and 1/64 seconds, but in no case was the cleavage rate greater than that of the controls — in fact, there was a slight delay when the dosage of 1/4 second was given.
Effect upon the Medium
When the mercury-argon discharge lamp is in operation, ozone in readily detectable quantities is produced by the action of the short ultra-violet (1849A) on the oxygen of the air. To determine whether this was dissolving in the medium and causing retarded development in the experiments reported above, sea water was irradiated for 64 and 256 seconds, then eggs were added and the development compared with the controls (5 experiments, 256 seconds; 10 experiments, 64 seconds). Unfertilized eggs, eggs 15 and 90 minutes after insemination and 2-celled stages were used. In all cases there were no signs of retardation, cleavage in all cases being comparable to the controls. The retarded cleavage of eggs irradiated with the mercury-argon tube
336
ARTHUR C. GIESE
is apparently due to the absorption of the ultra-violet radiation by the eggs and not to an effect upon the medium, and it is probably due practically entirely to radiation of X2537A, since the only lines of
TABLE II
Later development of eggs irradiated at various stages in development
|
Dos- |
Percent- |
No |
||||
|
age |
Irradiated at |
1 1 hours after |
age of |
27 hours after |
45 hours after |
of' |
|
in sec- |
lol. ,-tiiKi- ill (li-Vrlopnirnl |
insemination |
blastulie ab- |
insemination |
insemination |
eggs used |
|
onds |
normal |
|||||
|
1 niVrt ili/ed |
10r( non-motile |
0.0 |
All motile |
Normal gastrula? |
155 |
|
|
egg |
blastuUe |
|||||
|
15 inin. after |
4% non-motile |
().() |
All motile |
Normal gastruhe |
271 |
|
|
1 |
insemination |
blastulae |
||||
|
90 inin. after |
5% non-motile |
0.5 |
All motile |
Normal gastruhe |
198 |
|
|
insemination |
blastulae |
|||||
|
2 -ce 1 1 ed |
8% non-motile |
1.0 |
All motile |
Normal yastrulae, |
188 |
|
|
stage |
blastulae |
-mailer than con- |
||||
|
trols |
||||||
|
Unfertilized |
10% non-motile |
0.9 |
All motile |
Normal gastrulae, |
.U6 |
|
|
egg |
smaller |
|||||
|
15 min. after |
77' , non-motile |
0.0 |
All motile |
Normal gastruhne |
121 |
|
|
insemination |
||||||
|
4 |
90 min. after |
85( '( non-motile |
11.6 |
All motile |
Normal gastrulae |
104 |
|
insemination |
smaller than con- |
|||||
|
trols |
||||||
|
2 - c e 1 1 c cl |
99% non-motile |
7.6 |
All motile |
Normal gastruhe, |
144 |
|
|
stage |
few incompletely |
|||||
|
invaginated |
||||||
|
Unfertilized |
All non-motile |
0.0 |
Almost all |
Some abnormal; |
250 |
|
|
egg |
non-motile |
most smaller than |
||||
|
controls |
||||||
|
15 min. after |
All non-motile |
13.7 |
20f> motile |
10' ,' normal gas- |
257 |
|
|
insemination |
trula'. Others ab- |
|||||
|
normal moruhe |
||||||
|
16 |
90 min. alter |
All non-motile |
24.0 |
76cv motile |
87%gastrulate, |
175 |
|
inscMiim.tt ion |
small, somewhat |
|||||
|
abnormal |
||||||
|
2 -eel led |
All non-motile |
30.8 |
5% motile |
5% normal gastru- |
94 |
|
|
stage |
l.e, others abnormal |
|||||
|
morulae |
||||||
|
0 |
Control- |
M<ilile lil.i-lll- |
0.4 |
Motile M.i- |
Gastrulae |
232 |
|
u- |
tula; |
significant intensity in the mercury-argon tube spectrum are 3130, 3660, and several visible lines (4050, 4358, 5461, 5790A), all of which are relatively inert.
ULTRA-VIOLET RADIATION AND CLEAVAGE
337
Experiments with Pure Light of \2537A
To be certain that monochromatic light of X2537A is the effective agent, a number of experiments were tried with pure light of this wave- length obtained from the monochromator already described. The un- fertilized eggs were placed in a quartz cell and irradiated in the manner described in the previous paper (Giese and Leighton, 1935) usually for periods of 1, 4, 16, 64, 256 and 1,024 seconds. They were then trans-
Q.
3456
TIME SINCE INSEMINATION (HOURS)
8
FIG. 3. Retardation of cleavage, X 2537A; intensity, 9.74 ergs/mm.2/sec.
A. Control.
B. Sixty-four seconds exposure, extinction per egg: 2.91 ergs or 3.74 X 1011 quanta.
C. Two hundred and fifty-six seconds exposure, extinction per egg: 11.67 ergs or 1.50 X 1012 quanta.
D. One thousand and twenty-four seconds exposure, extinction per egg: 46.69 ergs or 6.00 X 1012 quanta.
ferred to watchglasses, inseminated and observed at hourly intervals until they had reached the 64-celled stage and then at 12-hour intervals until the gastrula stage. Seven series of experiments, the last three covering only the longer exposures, gave similar results and one series is plotted in Fig. 3.
It will be noted that retardation is evident when the dosage is
338
ARTHUR C. GIESE
large enough. Smaller doses produced no noticeable retardation or acceleration. The results are in general similar to those already obtained with the mercury-argon discharge tube.
Extinction Measurements
It would be interesting to know the amount of energy which must be absorbed to produce the above effects. By interposing a cell first empty then full of eggs between the thermopile and the light source one can determine the fraction of the light incident upon the cell which is transmitted, ///o. where 7 is the intensity of the transmitted light and 70 the intensity of the incident light. The fraction of the light extinguished, i.e., lost on passage through the cell can then be de- termined by subtracting the fraction transmitted from unity, 1 • 7/70. By determining the fraction of the area of the cell occupied by the eggs, Ac, one can determine the fraction of the light incident upon the eggs which is extinguished, 1 - 7/70 X l/Ac.
TABLE III
Data on extinction of light by sea urchin eggs
|
Ac, frac- |
||||||||||||
|
tion of |
1 — ///o for X's below: |
1 - 7//o X -T- for X's below: |
||||||||||
|
No |
area of |
Ae |
||||||||||
|
lixp. |
eggs |
cell bot- |
||||||||||
|
tom cov- |
||||||||||||
|
ered by |
||||||||||||
|
eggs |
2S37A |
1350A |
5461A |
5844A |
W04A |
2537A |
U50A |
S461A |
5844A |
6904A |
||
|
1 |
3320 |
0.49 |
0.51 |
— |
— |
— |
0.13 |
1.04 |
|
— |
— |
0.26 |
|
2 |
4431 |
0.65 |
0.63 |
0.38 |
0.27 |
0.23 |
0.14 |
0.97 |
0.58 |
0.42 |
0.35 |
0.22 |
|
3 |
3932 |
0.58 |
0.66 |
— |
0.34 |
0.26 |
0.14 |
1.14 |
— |
0.58 |
0.45 |
0.24 |
|
4 |
5964 |
0.88 |
— |
— |
— |
0.38 |
0.25 |
— |
— |
— |
0.43 |
0.28 |
Ac can be determined from the number of eggs, which can be counted, and the area of the egg effective in extinguishing lighl, which is approximately the area of a circle of the diameter of the egg. The average <li. under of the egg used was the mean of the diameters o! 50 eggs taken at random, 77.2 ± 2.9 /z (two diameters were measured since almosi one-half of the eggs were slightly ellipsoidal). A circle of i his diameter has an area of 4,681 fj,-. About 6,765 eggs would be necessary to completely rover the bottom of the cell used (area, 31 .67 X 106 /r) u'ith a layer one egg diameter thick. Ac is the number of eggs counted in a given experiment divided by 6,765.
The transmission of light of various wave-lengths by eggs was determined, the method used being similar in practically all respects lo thai described for Paramecium (Giese and Leigh ion, 1()35), and (he number of eggs was counted in each experiment. The experimental and derived data are recorded in Table III.
ULTRA-VIOLET RADIATION AND CLEAVAGE
It is clear that practically all the light of X2537A incident upon the eggs is extinguished, while at the longer wave-lengths a much smaller proportion is so lost. Part of the light extinguished is truly absorbed, part is lost by scattering from the surface of the egg and from the surfaces of small particles within the egg. Unfortunately it is very difficult to measure the scattering and it is impossible to obtain an approximation of the scattering as was done with Paramecium since unlike the latter the eggs absorb in the visible part of the spectrum as is quite obvious from the data in Table III. It is probable that as for Paramecium a considerable proportion of the light extinguished is actually scattered, possibly as much as 50 per cent of the total. How- ever, the extinction measurements at least give the order of magnitude of the energy involved. The extinction by the eggs for the various experiments in Fig. 3 has been determined and the data have been added to that figure.
DISCUSSION
The data of Fig. 3 are interesting because they give an idea of the number of quanta which must be absorbed to produce an effect. Thus eggs extinguishing between 1.57 X 101J and 2.54 X 1010 quanta, and probably absorbing about half this quantity, were not visibly affected and cleaved comparably to controls, only after a dosage of about 3.74 X 1011 quanta per egg was the rate of cleavage definitely retarded. Only doses short of those producing cytolysis stop cleavage for eggs extinguishing 1.24 X 1013 quanta went on developing as far as the 8- celled stage in many cases. Were one to assume that the average molecule in the egg protoplasm had a mass of the order of magnitude of the mass of the egg-albumin molecule, one would find the egg to possess some 2.5 X 1012 molecules. The actual number is probably much larger, but the figure indicates that a fair proportion of the molecules are affected or that certain molecules have absorbed many quanta before an effect is evident. The data are also indicative of the high power of recovery from injury possessed by the egg protoplasm.
From the data presented one may conclude that for the wave-length and the dosage series used, which covers the range usually employed in similar experiments, there is no evidence of acceleration of cell division. Following large doses of radiation there is retardation; following smaller doses the rate of cleavage is not noticeably different from controls.
These results do not, however, exclude the possibility of a stimu- lative effect of doses of ultra-violet light much weaker than here
340 ARTHUR C. GIESE
employed. The radiations claimed by the Gurwitsch school of mito- genetic rays to be short ultra-violet and to be effective in increasing mitoses are postulated to be an entirely different order of magnitude from the radiations here used, in fact so weak as to defy physical detection. No attempt is here made to throw light upon this complex problem (see Rahn, 1936).
SUMMARY
1. There is a threshold dosage between 10l(1 and 10" quanta below which radiation of X2537A produces no observable change in the rate of cleavage. Beyond this threshold the degree of retardation increases with the dose.
2. Many of the retarded eggs develop normally, but are slower in reaching a given stage; others continue developing for only a short lime, the degree of differentiation reached being inversely proportional to the dosage.
3. Unfertilized eggs, eggs 15 and 90 minutes after insemination, and eggs in the first cleavage do not exhibit strikingly different sus- ceptibilities to the rays, although the later stages appear to be some- what more susceptible.
4. The quantity of radiant energy which the eggs can absorb before being affected is quite large, as indicated by the extinction measure- ments reported, and serves as a rough measure of the power of repair of the egg protoplasm.
5. A series of dosages from a dose which cytolyzes to one which has no retarding effect upon cleavage with light of X2537A failed to induce artificial parthenogenesis.
6. No evidences were obtained over the dosage series investigated for acceleration of the rate of cleavage.
BIBLIOGRAPHY
ALP.VIOV, \Y. \V., ANI)(). K. NAM M KOVA, 1933. The influence of different <]u, unities
of ultra-violet radiation on the division rate in Paramecium. Protoplasma,
18: 2X1. BOVIE, W. '!'., AND I ). M. HUGHES, 1918. The effects of quart/ ultra-violet light upon
the rate of division of I'aramecium caudal 11111. Jour. Mcd. Res., 39: 223. CHASE, II. Y., I'M". The effect of ultra-violet light upon early development in
eggs of Urechis caupo. Biol. Bull., 72: 377. COBLENTZ, W. W., 1931. Sources of radiation and their physical characteristics:
cold red ray and cold ultra-violet ray lamps. Jour. Am. Med. Assn.,
97: 1965. COBLENTZ, W. W., AND II. R. FULTON, 1924. A radiometric investigation of the
germicidal action of ultra-violet radiation. U. .V. Bur. Stand. Sci. Paper
No. 495, 19: 641. GATES, 1 . I.., 1929. A study of the bactericidal action of ultra violet light. I. The
reaction to monochromatic radiations. Jour. Gen. Pliyxiol., 13: 231.
ULTRA-VIOLET RADIATION AND CLEAVAGE 341
GIESE, A. C., AND P. A. LEIGHTON, 1935. Quantitative studies on the photolethal effects of quart/ ultra-violet radiation upon Paramecium. Jour. den. Physiol., 18: 557.
HERTEL, E., 1905. Ueber die Einwirkung von Lichtstrahlen auf den Zellteil- ungspro/ess. Zeitschr. allgem. Physiol., 5: 535.
HINRICHS, M. A., 1928. Ultra-violet radiation and division in Paramecium cauda- tum. Physiol. Zool., 1: 394.
HUGHES, D. M., AND \V. T. BOVIE, 1918. The effect of fluorite ultra-violet light on the rate of division of Paramecium caudatum. Jour. Med. Res., 39: 233.
HUTCHINSON, A. H., AND M. R. AsHTON, 1929. The specific effect of monochromatic light on the growth of Paramecium. Can. Jour. Res., 1: 292.
LEIGHTON, P. A., AND W. G. LEIGHTON, 1932. Some remarks on the use of thermo- piles for the absolute measurement of radiant energy. Jour. Phys. Chem., 36: 1882.
LEIGHTON, W. G., AND P. A. LEIGHTON, 1935. Visual demonstration of the evapo- ration of mercury. Jour. Chem. Ed., 12: 139.
OSTER, R. H., 1934. Results of irradiating Saccharomyces with monochromatic ultra-violet light. I. Morphological and respiratory changes. Jour. Gen. Physiol., 18: 71.
RAHN, O., 1936. Invisible radiations of organisms. Protoplasma Monographien, vol. 9, Borntraeger, Berlin.
THE HISTOLOGY OF THE RETRACTOR MUSCLES OF THYONE BRIAREUS LESUEUR
MAGNUS OLSON
(From the Marine Biological Laboratory, Woods Hole and tlie I Department of Zoology, University of Minnesota)
The histology of invertebrate muscle has received relatively little investigation and there exist at the present time marked gaps in our knowledge of this subject. A more critical study of invertebrate muscle might possibly yield clues of interesting phylogenetic signifi- cance. Recent research in the field of holothurian muscle physiology has also made desirable a more complete histological study of these muscles.
The only extensive paper on the histology of holothurian muscle (Hall, 1927, on Cucumaria) completely misinterprets the essential structure and has been responsible for a number of erroneous conclu- sions by workers in muscle physiology.
This paper concerns itself with the more salient features of the histology of Thy one retractor muscle. The interpretations are op- posed to the conclusions of Hall on the strikingly similar muscles of Cucumaria.
MAT K RIALS AND MKTIIODS
The retractor muscles of Thyone occur in five main bands which run from the lantern to points of attachment on the five longitudinal inux Ics of the body. The individual bands may be single or broken up into a number of divisions. They are extremely extensible and in adult specimens may vary in length from about 5 mm. in the contracted condition to 5 or 6 cm. in normal extension. Except for the two regions of attachment the muscles are free of the tissue and are easily removed.
The muscles \\ere fixed in varying degrees of contraction in Bouin's or Helly's fixing fluids. In order to obtain normal extension it was necessary to immerse them for some time prior to fixation in 0.3 molar magnesium chloride solution. The muscles were dehydrated in alcohol, cleared in xylene, and imbedded in paraffin. Sections were cut at 5 and 10 micra.
I leidenhain's iron ha-matoxylin proved to be the most useful stain for the muscle libers. Mallory's aniline blue and I )oininici's
342
HISTOLOGY RETRACTOR MUSCLES OFTHYONE
stain \vere most favorable for the details of the connective tissue network.
Attempts were made to macerate the muscles in Bouin's fluid for periods up to three weeks and in varying strengths of nitric acid. These techniques, however, did not prove to be very satisfactory.
OBSERVATIONS
Microscopic examination of the fresh muscle reveals extremely long fibers grouped into indistinct bundles. In cross-sections of fixed and stained muscles this bundle arrangement is clearly discernible (Figs. 1 and 4). The inner portion of the muscle consists of hemi- cylindrical bundles containing from two to fifteen fibers which tend to be arranged about the periphery of each bundle. The outer portion of the muscle exhibits no definite bundles. Longitudinal sections reveal that the bundles do not maintain the same complement of fibers throughout their length. Fibers pass from one to the other of adjacent bundles and in the periphery of the muscle there is no definite association of fibers.
The spaces between the bundles and the fibers of the bundles are occupied by a delicate reticulum of connective tissue which is markedly condensed about each bundle (Figs. 1 and 4). The connective tissue appears to have the nature of a lattice-like framework with the thickest fibers running perpendicular to the muscle fibers. In sections of contracted muscle the elongated connective tissue nuclei are also seen to lie with their long axes in this direction. The connective tissue fibers are especially coarse and dense in the periphery of the muscle.
Connective tissue and tissue space constitute a relatively large proportion of the muscle in contrast to mammalian smooth muscle and some invertebrate muscle, e.g. retractor muscle of Phascolosoma. Steinbach (1937), in his studies on potassium and chloride in Thyone muscle, has estimated the volumes occupied by the muscle elements and tissue space. According to Steinbach, chloride diffuses freely into 43 per cent of the muscle — a figure which by his interpretation corresponds rather closely to the extrafibrillar space (tissue space and sarcoplasm). However, since Steinbach tentatively accepted the histological conclusions of Hall, his results may be reinterpreted. The structures which Steinbach, following Hall, interpreted as fibrils are actually fibers. Those structures interpreted as fibers are bundles of fibers. Thus it is found that chloride space corresponds closely to tissue space (47 per cent of Thyone muscle) as has commonly been believed (see Fenn, 1936), and does not include the sarcoplasm of the muscle fibers.
344 MAGNUS OLSON
Fischer (1937) found that Thyone retractor muscle exhibited a birefringence of 60 per cent of that of Phascolosoma retractor muscle. (Historically, Phascolosoma retractor muscle exhibits densely packed elements.) Since birefringence may be presumed to be associated with contractile elements, Fischer's figure suggests that these elements are fewer in number than in Phascolosoma muscle, and it may be further interpreted as lending support to the figure given above for tissue and chloride space in Thyone muscle.
As far as it has been possible to determine, the fibers are uni- nucleate. The nuclei occupy lateral positions on the fibers (Figs. 1 and 3), only rarely are they found within them (Fig. 4). In the contracted fiber the nuclei are rounded and are usually found to occupy a typical lateral position on the fiber. With the proper staining technique a layer of faintly staining cytoplasm may be seen to surround the nuclei. Often a thin strand of this cytoplasm forms the only connection between the nucleus and the muscle fiber. In the extended muscle the nuclei are elongated in the axes of the fibers and lie closely applied to them.
The appearance of the fibers in cross-section is somewhat variable. In the contracted condition they tend to be circular or semicircular (Fig. 1). Occasional fibers are polygonal. The stretched fibers may be circular or oval in transverse section, but are more often flattened or angular (Fig. 4). The fiber diameter varies from 5-10 ju in con- traction to 2-4 n in extension.
Teased preparations of muscle reveal the spindle-shaped character of the fibers. The tips of the fibers are long and tapering in extension (Fig. 5), bluntly rounded in contraction.
It was extremely difficult to determine the lengths of the muscle fibers since it was almost impossible to isolate these elements. The closely investing connective tissue binds the fibers securely together and all attempts at maceration and teasing were largely unsuccessful. Careful study of the best of these preparations, however, revealed that the fibers do not extend the whole length of the muscle. At
Explanation of Figures
All drawings were made with the aid of a camera lucida at a magnification of X 745.
FIG. 1. Cross-sections of two bundles of fibers from a contracted muscle.
FIG. 2. Portion of a longitudinal section of an extended muscle.
FIG. 3. Portion of a longitudinal section of a contracted muscle.
FIG. 4. Cross-sections of two bundles of fibers from an extended muscle.
FIG. 5. Fragment of a teased muscle.
HISTOLOGY RETRACTOR MUSCLES OFTHYONE
345
1
>0
i*
, (
(
.1
I J/
,
0
PLATE I
\ \ \\\
5
346 MAGXUS OLSOX
various points along a teased muscle natural unbroken tapering fiber terminations could be found (Fi.u- 5).
DISCUSSION
The contention of Hall (1927) that Citcumaria muscle consists of large fibers containing a feu -iant fibrils forms an unten.ible theory. From Hall's figure^ it is apparent that the histological picture of Cncnmarin mu-cle is almost identical with that of Thyouc. In Thyone muscle. however, the nucleated condition of the so-called fibrils is clearly evident. It must be concluded, therefore, th.it these are mn-rle iil>ei> and not fibrils. It is true that the nuclei often have a tenuous attachment to the fibers, but this attachment is nevertheless definite, and, as previously stated, nuclei are occasionally found within the libers. The free nuclei within the bundles, which Hall interpreted ,i- muscle nuclei, are clearly constituents of the connective tissue. I nrther, the failure of the peripheral elements to be arranged into bundles (muscle fibers of Hall) does not support Hall's conclusion that these are fibrils.
The conception that the muscle libers are long spindle-shaped elements is supported by the studies of Gerould (18%) on Candina and Jordan (1914) on Holothuria and Slichopiis. According to ( ierould. muscle fiber nuclei in C/iinliiKt occupy a lateral position similar to those in Thyone.
du Buy (1936), studying the sensitivity of Thyone muscle to certain drugs, found that these muscles exhibited a physiological condition intermediate between vertebrate smooth and vertebrate -iriated muscle. Microscopically they resemble vertebrate smooth muscle. No histological similarity with vertebrate striated muscle Can be detected with the possible exception of their extreme length. The libers are homogeneous, devoid of striations or fibrillar structure.
SUMMARY
The retractor muscles of Thyone consist of long spindle-shaped smooth nin-cle libers. The libers appear to be uninncleate. The nuclei usually occupy .1 Literal position and are attached to the liber by a delic.iie |.(yer of cytoplasm. < )ccasional nuclei are found com- pletely within the liber-.
In the ( enter o| the nm-cle the libers occur in bundle- ol two to fifteen. V> Mich organi/ation is exhibited in the periphery of the muscle.
The fibers do not remain associated with a single bundle throughout their length but pass freely from one bundle to the other.
HISTOLOGY RETRACTOR MUSCLES OFTHYONE 347
Measurements of fiber length were not obtained, but it is evident from study of teased preparations that they do not extend the full length of the muscle. Fiber diameter varies from 5-10 micra in contraction to 2-4 micra in extension.
The connective tissue forms a lattice-like network which is concen- trated about each bundle and in the periphery of the muscle. The heaviest fibers of the reticulum tend to run perpendicular to the axes of the muscle fibers.
LITERATURE CITED
DU BUY, H. G., 1936. Separation of the conducting and contractile elements in
the retractor muscle of Thyone briareus. Biol. Bull., 71: 408. FENN, W. O., 1936. Electrolytes in muscle. Physiol. Rev., 16: 450. FISCHER, E., 1937. The influence of length, tension and tone upon the birefringence
of smooth muscles (Phascolosoma and Thyone). Biol. Bull., 73: 363. GEROULD, J. H., 1896. The anatomy and histology of Caudina arenata Gould.
Harvard Bull. Mus. Comp. Zool., 29: 123. HALL, A. R., 1927. Histology of the retractor muscle of Cucumaria miniata.
Publ. Puget Sound Biol. Sta., 5: 205.
JORDAN, H., 1914. Uber "reflexarme" Tiere. IV. Zool. Jahrb., 34: 365. STEINBACH, H. B., 1937. Potassium and chloride in Thyone muscle. J. Cell, and
Comp. Physiol., 9: 429.
B & L TYPE G MICROSCOPES
QUALIFY FOR EXACTING
RESEARCH WORK
There are 8 different models of the B & L Type G Microscopes available, meeting the most exacting requirements for every phase of research microscopy. All possess the accuracy, rigidity and precision so essential to research work.
A wide range of accessories enable the adaptation of even the simplest and least expensive model in .iilv.mred work. Vertical or inclined, monocular or binocular body tubes; plain or mechanical stages, rectangular or circular; simple or centering substages, for use with every kind of illuminating apparatus; and a wide variety of eyepieces and objectives for every type of work are included among their features.
Write for Catalog D-I29 which describes B £ LTypeG Microscopes adapted to every specific research need. Address Bausch & Lomb Optical Co., 652 St. Paul Street, Rochester, N. Y.
BAUSCH &. LOMB
. . . . wt MAKI OUR OWN GLASS TO
INSURE STANDARDIZED PRODUCTION
f OR r OUR GLASSES INSIST ON B A L ORTHOGON LENSES AND B » L FRAMES . . .
CLAFF RECOVERY DISH
See article on page 178 of this issue by George W. Kidder and C. Lloyd Claff, "Cytological Investi- gations of Colpoda cucullus."
No. A-1470 Each $.35 Dozen $3.5O Recovery book supplied with each dozen.
CLAY-ADAMS CO., Inc. - 25 E. 26th St. - New York
Vol. LXXIV, No. 3 June, 1938
THE
BIOLOGICAL BULLETIN
PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY
A SIMPLE TELEOST KIDNEY IN THE GENUS
CYCLOTHONE
BRADFORD B. OWEN
(From the Biological Laboratories, Harvard University')
INTRODUCTION
Because it is the unit of structure and function in all vertebrate kidneys, the renal tubule contains the answer to many questions con- cerning the production of urine. Yet any theory which attempts to assign the various steps in this process to separate regions of the tubule must take into account the surprisingly wide degree of variation in tubule structure among the vertebrate groups. The relation of the presence or absence of the glomerulus to urine production and to the osmotic balance between the organism and different sorts of environ- ment has been clearly outlined by H. W. Smith (1932) and others. The control of water content and salt concentration depends not only upon the action of the glomerulus, but also upon resorption and excre- tion by other tubule portions, such as the loop of Henle in mammals, and upon entirely extra-renal factors, such as the excretion of urea and salts by the gills of fishes.
The problem of further localizing kidney functions among the re- maining segmental differentiations in the tubule proves to be more difficult. The fish kidney has been the subject of several cytological and histophysiological investigations contributing to this question, but confusion has resulted from the extreme diversity of kidney types in this group. Very few are found to contain all the segments of the typical vertebrate nephron, and Marshall (1934) has concluded that only the proximal convolute is common to all. There are as yet few clues to the benefits conferred by the other segments upon those fish which possess them. Among the forms which are able to adapt to sudden changes from fresh to salt water and back again, Grafflin (1937a) finds a lack of special development in any particular segment in relation to this remarkable ability, and concludes that it must be due to entirely extra-renal factors. Yet he does find indications of homolo- gies in proximal convolute structure among those few which have been
349
350
BRADFORD B. OWEN
studied cytologically (see below). It is obvious that there is a need for observations on more forms, especially because of the wide diversity of tubule types in the few already examined. The present study was undertaken with this fact in mind, and it was thought that some of the little known deep >eu fish might present interesting modifications. The inaccessibility of their habitat, and the numerous strange adapta- tions of body structure which they have produced there, might be taken as indications of an early arrival in the sea, which would have given plenty of time for kidney structure to reach the aglomerular condition considered most logical for marine forms. Actually, the few genera so far sectioned all show glomeruli. One of these, however, shows a strikingly .simple type of kidney, and forms the subject of this paper.
TABLE I Data on the seven Cyclothone sectioned for study of kidney in situ
|
Series Number |
Length in mm. |
Depth taken |
Fixation |
Serial sections |
Portion of kidney included |
|
1 |
55 |
— |
Formalin and sea water |
10 M |
Complete |
|
2 |
40 |
— |
Modified Held's |
10 n |
Complete |
|
3 |
56 |
0-2000 m. |
Bouin's |
SM |
Complete |
|
4 |
28 |
600 m. |
Bouin's |
SM |
Complete |
|
5 |
26 |
600 m. |
Bouin's |
SM |
Complete |
|
6 |
40 |
— |
Modified Held's |
10 AI |
Posterior, from proximal convo- lute 2nd part to end. |
|
7 |
45 |
— |
Formalin and sea water |
10 AI |
Anterior, as far as proximal convolute 2nd part. |
MATERIAL AND METHODS
T!i<- fish used in this investigation is a small deep sea form belonging t<> tin- genus Cyclothone. It is long and very slender, without eyes, and supplied with numerous minute photophores. The specimens range from _'<i to 5f> nun. in length, with corresponding diameters of 3 to 6 mm. respectively. They were obtained in the vicinity of Bermuda, mostly in open-net hauls at depths of 0 to 2,000 m. Some were caught in closing-net hauls at 600 meters and other depths. The collections were made from the ketch Atlantis of the Woods Hole Oceanographic Institution during the summers of 1936 and 1937, and specimens for
TELEOST KIDNEY IN CYCLOTHONE 351
the present work were obtained through the courtesy of Dr. J. H. \Yelsh of Harvard University.
The catch from each haul was usually preserved en masse in formalin and sea water. The extremely delicate body of the fish was shorn of much of the integument by crowding and rough treatment in the net, and was further twisted and distorted, and fixed in this state. Thus it was difficult to find a specimen in a straight condition suitable for sec- tioning. A few had been fixed separately in Bouin's or a modified Held's, but were in no better condition. None had been cut open, so that fixation was relatively poor. A group of those showing signs of better than average fixation plus a minimum of distortion wras selected for a study of the kidney in situ by means of serial sections of the whole animal. In five of these, enough of the bodies were sectioned to include the entire urinary systems, which are reproduced approxi- mately to scale (Figs. 1-5). A smaller portion of two others also was sectioned. All were stained with Harris hsematoxylin and eosin. The data concerning the seven are given in Table I.
OBSERVATIONS
The Cyclothone kidney consists of a single pair of renal tubules extending parallel to the long axis of the body. They begin as two large, closely approximated Bowman's capsules, located dorsally at about the level of the last gill arches, and run a practically straight course side by side until they unite to form the bladder anterior to the anus. An investment of hemopoietic tissue surrounds the anterior half of the tubules, and just before entering the bladder they pass through a mass of glandular tissue which is probably the organ of Stannius.
Determination of the exact amount of regional differentiation in this nephron is difficult in the material available at present. There is a glomerulus, a neck segment, and a proximal convoluted segment of two portions. These four regions are marked off by histologically abrupt transitions. The second portion of the proximal convolute is followed by a thin intermediate segment, and this by a terminal segment somewhat like the distal convolute described in many tubules. These last two, however, are not set off by abrupt shifts in cell types. It is necessary to describe a long transitional region before the intermediate segment and, although the terminal segment appears rather suddenly, there is not a sharp histological transition from the intermediate segment.
Glomerulus and Neck
The capsule of Bowman is large, with a well-vascularized glomerular tuft which lacks the central avascular core observed in bird and reptile
352 BRADFORD B. OWEN
nephron by Yilter (1935). There is a large amount of intracapsular space, but it is not filled with coaguluni. and does not resemble the degenerative enlargements or outpouchings seen by Grafflin (1929) in the goosefish capsule.
The two capsules are separated by a thin septum (Fig. 6) which becomes indistinct as it passes between the capillary tufts, so that these latter structures appear to form a single unit. It was not determined whether there wa- a M ]>arate blood supply for each side of the tuft.
The short neck segments of the two tubules arise ventrolaterally from the glomerular capsules, and there is an almost immediate transi- tion to the cells of the proximal convolute (Fig. 6). The neck cells bear cilia, but these are relatively few and very fine, requiring oil im- mersion for identification. The segment therefore does not give a pic- ture comparable to the ciliated neck of other nephrons, such as that described by Guyton (1935) in the lungfish. The neck cells are of a low cuboidal type, 5 to 8/x, in height. Nuclei are small, irregular in shape, and stain densely. The cytoplasm also takes a hamiatoxylin stain. This is a short segment in all specimens, usually less than 100 p. The diameter is small, but the lumen relatively large, so that there seems no question as to the functional nature of the glomerulus.
First Portion of Proximal Convolute
The cells of this region appear abruptly (Fig. 6). In the first few sections, the tubule wall shows a peculiar tufted structure due to the fact that the cell height varies from 12 to 50 p. (Figs. 7 and 8). Very sooiij however, the cell height becomes comparatively constant (Fig. 9). Two of the specimens fail to show this irregularity.
In one member of the series (No. 7) the cells of the "tufted'1 -rion just described are filled with very large granules (Fig. 10) which crowd the nuclei against the basement membrane. These granules are -pherical in shape and appear homogeneous in composition. Most of them stain a brilliant red with eosin. although a few remain colorless. • Mli' r -j.rcimrns show the granules to a lesser extent, always smaller and fewer in number.
I-'n.-. 1 tn 5. Diagrammatic reconstruction of kidneys of five Cyclothone specimens. X 10. Only one tubule is shown; the other joins it at beginning of bladder. l~pp' r Mippln] areas, lu-mopoictic tissue; lower stippled areas, organ of Stanniu-. G, ;ilus : A", ncrk segment; P, first portion of proximal con-
volute; P=, seo'ti'l portion of proximal convolute; TRANS., transitional zone; LV., intermediate segment; '/'. t ••ment; B, bladder. Arrows indicate level
of appearance of first nliatnl oils. X umbers correspond to those of the scries described in Table 1.
TELEOST KIDNEY IN CYCLOTHONE
353
TRANS.
IN.
TRANS.
IN
B
V
TRANS
IN.
B
TRANS.
IN.
B
[ TRANS
IN.
FIGS. 1-5
354 BRADFORD B. OWEN
The apical cytoplasm is always more or less granular, while that below the nucleus shows prominent striations, with fibrilke arranged parallel to the long axis of the cell and closely packed together. The cytoplasm takes a deep eosin stain. There is a uniform brush border, rather coarsely fibrillar. and 5 to # p. in height.
Nuclei are oval, 4 to 5 p. in diameter, and heavily stained. Their position within the cell varies widely. In some specimens, they are located below the center of the cell in the first few sections, but later become central or even apical. In one series the nuclei are in the basal half of the cell throughout, while in another the position is always in the apical half (Figs. 7 and 8). The tubule enlarges greatly toward the end of this region, and convolutions appear. These always have the same simple structure, a double fold (Figs. 1 to 5).
Second Portion of Proximal Convolute
At this point there is an abrupt change (Figs. 11 and 12) to a cell type with lighter staining cytoplasm, larger nucleus, and high, irregular brush border. The cytoplasm shows very few granules, but is still stri- ated, with fibrillse loosely packed so that the cell appears less dense. The brush border becomes exceedingly irregular in form (Figs. 12 and 13) sometimes extending out into the lumen for a distance of 25 p.. Its structure is finely granular and sometimes vacuolated, without stri- ations. Some portions have a droplet-like appearance as if about to be pinched off into the lumen, which suggests that these cells are engaged in some sort of apocrine secretion. In later sections, the border be- comes lower and more regular, sometimes regaining a striated ap- pearance.
The nuclei are larger than in the first portion and stain less heavily (Figs. 11 and 12). They are usually just above the center of the cell.
I:i<;. 6. Section of kidney at level of glomeruli, showing the close approxi- mation (if these bodies. Right tubule (above) sectioned at level of transition from neck (right) to Pt (left). Other tubule (below) shows scattered P, cells among those «'f neck. Passage from glomerulus into neck is in an earlier section. X 180.
I-i'.. 7. Tufted region in early part of P,; in this specimen the nuclei are basal and there i> a large quantity of granular cytoplasm between them and the lumen. X 255.
FIG. 8. Same region of another specimen. Here the nuclei are apical, and there is much less of the granular cytoplasm. X 240.
FIG. 9. Central part of /*,. The tubule wall is even and the brush border regular. X 130
FIG. 10. Cells of /',, showing remarkable number of large granules. These occur in only one specimen; see text. X 535.
FIG. 11. Section passing through region of transition from /J, (cells of upper right half) to /',. (lower left). X 280.
TELKOST KIDNEY IN CYCLOTHONE
355
' '<
m
&.
F>'<
tf
s
•;
r.
X1
•«
•»
>
.,' *f?
7
\ *-<IS.-?B«fc«
;^^,
^:^' ^ V ^.^ . *
-V'vjv- '
*«.*tf ?V^ •**"> ^ i ^»
1^!' •..-^-«i "a • \.\ 5.^»
FIGS. 6-11
356 p, RAD i- OKI) i!. o\\T.X
Kre<[uent double nuclei are seen. oriented in a line parallel t<i the vertical axis of the cell ( Fig. 13).
The cells are wider than in the first portion, hut only slightly higher. ranging from 15 to Jo,, in height near the- transition from the first half of the segment. There i> a steady decline in height of cell and brush border both, throughout this
Intermediate Segment
When tlu- brush bonier is cntirelv gone and tin- cells are reduced to a cuboidal shape without ])roininent >triation> in the cvtoplasin. a tvpical intermediate segment has been re-ached. This is >e]iarated from the ])ro.\imal convolute by a long transitional /one in place of the abrupt change noted h-tween the other segments. Two factors make it im- possible to eliminate this /.one. In the first place, the changes in cell height and nature of bru>h border take place gradually. The nucleus also slowly becomes smaller and more densely stained. In the second place, the cells on the mi-dial side of each tubule retain the tall brush border, greater cell height, and tibrillar cytoplasm long after the cells ot the rest ot the epithelium are of the low cuboidal type of the inter- mediate segment. without brush border or cytoplasmic .striations.
I lie tubule m the typical intermediate segment is verv small and lies on the dorsal wall of the body cavity without any investment of liemo- pojetic tissue. It is usually less than 5<)/t in diameter, and the cells are otten much wider than high. The cytoplasm is homogeneous, dense, and rather heavily stained with eosin. Although the cell border is typically plain, there are trei|uentlv various sorts of bleb-like protrusions on the cell surfaces which are verv much like a brush border.
FlG. \1. Same region under higher magnification. \i>U- abruptness ,,f traiisi- ti'.n: /' cells ire mi riyht. X 'oil.
l-'n.. l.i. 'I'ypiral lii.yh linish l)onli-r of /',. It is highest un the medial wall of each mimic. X 240.
FlC. 14. Who] i- section; the kidney tuhules lie just above the oesophagus
l)m-sal side uppermost. Tlu- heart max he seen veiitrally. • ''. 15. r.r.niiuiiu.u nf (ransitioiial XDIH-. The ciiitlu-liuin is 1"\\, hnish border i- l.i-L'iiiuiiii' t«> disajiprar (upper left > and ciliated cells are o minimi. 365
I-'K.. I1'. Inteniu-di.iti- segment. V'erj l"\v epithelium, no linish Itorder, cilia ininir i Mi t< mall size of lulmle. X 720.
l;n.. 17. Portion of tulmle wall (/'...) to sho\\ a t\ |>ical isolated ciliated cell. X (.
Fit;. IS. Terminal -eL-ment. Tin- strong siiu^estii,n of a brush horder shown li- • most ol :l;e oilier s])erinu us. - 645.
Fi«.. I1'. Stanniu-. • <\ uan, with kidney tubule at either end of glandular mass. X 90.
TKLEOST KIDNEY IN CYCLOTHONE
357
(2
f fa
V. *<
- '3
^•tl
FIGS. 12-19
358 BRADFORD B. ONVEX
Cilia are fairlv numerous, but are so extremely fine in texture that they never form a conspicuous feature of the tubule topography, and frequently require high power to be -mi at all. Isolated ciliated cells first occur early in the second portion of the proximal convolute i Fii 1 to 5). They are conspicuous as small cells without a brush border and with heavily stained nuclei, wedged between the other cell- of the tubule wall (Fig. 17). The several cilia from each such cell are fused into one strand, only occasionally showing the individual elements. Rare at first. these cells become progressively more numerous up to the end of the intermediate segment, where1 then- may be several in one cross-section (Fig. 15). In both structure' and distribution they re- semble the isolated ciliated cells in the second portion of the proximal convolute of the lungfish ((iuyton.
As the tubule enters the organ of Stannius (Fig. I'M the cilia disappear, and there is a certain amount of restoration of the glandular type of epithelium. It shows an increase in thickness in all cases, usu- allv about 20 per cent, but in two specimens the cells arc- doubled in height, and regain the striated cytoplasm. In one of these there is also a marked accumulation of basophilic granules in the apical region of the cell i Fig. IS i. The- cell mar-ins arc uneven and sometimes tin- rough- lie-- suggests a brush border (Fig. IS). In certain regions, however, the luminal border is smooth, while some of the nephroiis show no sign of this surface modification at all, so that it is assumed this is not a brush border segment.
The transition to the terminal .segment takes place within the distance covered by five or M'\ sections, in one specimen, yet there is always a gradation of one cell type into the other, without an abrupt change.
It seems sate to conclude that then- is no collecting segment, for the glandular epithelium frequently lasts unchanged to the point of union of the tubules at the bladder. Some tubules show a gradual return toward the structure of the intermediate' segment, but the cells always remain somewhat taller.
Convolutions appear just befoie the start of the bladder. They are much les, regular than the first set described above'. The' large spevi- me-ns. which are females, show only a slight bending of the tubule-, while' in the small ones ( Figs. 4 and 5 ) it may go through several tor- tuous loops. h is interesting to note' that the'se smaller individuals are' niale-s, a fact which suggests that the- extra convolutions might hear some relation to the "sexual segment" observed in certain vertebrate1
TELEOST KIDNEY IN CYCLOTHONE
male kidney tubules. However, no connection was observed between the testis and the kidney in either of the two males sectioned.
In the midst of these convolutions, the tubule leaves the organ of Stannius and descends ventrally to the bladder. This organ is lined with a low cuboidal epithelium, in which the curved tops of the cells give a characteristic biscuit-like appearance. A very thin coat of smooth muscle completes the structure. An external opening is found behind that of the digestive tube, and close to that of the reproductive tract, at least in the case of the female. Neither of the male specimens shows an open duct leading from the testis, although a very thin strand of tissue can be traced back from this organ toward the region of the anus.
DISCUSSION
All the Cyclothone kidneys so far sectioned show a high degree of similarity between corresponding parts of the tubules from the glomer- ulus on through most of the second portion of the proximal convolute. Beyond this region differences are encountered in the nature of the epi- thelium, making it difficult to find characters common to all, and next to impossible to draw comparisons with other tubule types.
The confusion with regard to structure in this region is not entirely due to an unusual tendency toward individual variations. The proximal half of the tubule shows decidedly better fixation of important cellular features, probably because penetration was much faster there in the region of the gills than it wras posteriorly where the kidney is surrounded by the thick muscle layers of the body wall (Fig. 19).
Audige (1910) states that the typical teleost kidney may consist of three parts, the head, middle, and hind kidney. The head kidney in young fish contains two large glomeruli derived from the pronephros. These are rarely seen in adults, although Price (1910) describes one in BdcUostoma stoiiti. In Cyclothone there is a lack of any opening into the ccelom such as would be expected with a pronephric tubule, and in addition the glomerulus is not external. Furthermore, there is a direct passage toward a urinary bladder along a tubule wrhose epithelium is characteristic of the renal tubules of typical vertebrate mesonephroi. Nevertheless, the extreme anterior location of these two bodies suggests that they may be modified pronephric glomeruli, associated posteriorly with a pronephric duct which has undergone modification into a typical secretory tubule.
There is a remarkable variation in tubule sizes among the seven individuals sectioned. In the typical vertebrate kidney all tubules are identical, but their number varies with the size of the kidney and of the
360
BRADFORD B. OWEX
animal containing it. Xash (1931) has pointed out the constancy of the ratio between tubule number and body surface in several fish. In the present example, however, the tubule number is constant. The amount of kidney tissue is nevertheless adjusted to the size of the ani- mal, simply by a change in size and length of the tubules themselves. The size of the fish from which these tubules are taken is given in Table I.
Even the smallest of the five illustrated here is a rather large tubule in comparison with those of other fish. Table II lists the approximate measurements of the largest and smallest tubules of the Cyclothonc series (Figs. 1 and 5). These may be compared with tubule sizes reported by Xash (1931) for a large number of species. Tubule lengths he found usually between 1 and 9 mm.; diameters ranged upwards from 25 p. Glomeruli were commonly between 35 and 95 p. But
TABLE II
Approximate measures of largest and smallest tubules of Cydothone series
|
Segment diameters (M) |
Total length |
||||
|
Glomerulus |
Prox. conv. |
Interm. |
Terminal |
||
|
Largest (No. 1) |
250 |
200 |
50 |
60 |
14.2 mm. |
|
Smallest (No. 5) |
130 |
100 |
30 |
50 |
5.6 mm. |
isolated examples in his list far exceed this range: the hagfish glomeru- lus is 500 p. in diameter, and the nephron of the skate attains a length of 19 mm. Therefore the Cyclothonc nephron conies well within the limits observed by Xash. In only one specimen is the tubule diameter really exceptional ; the proximal convolute of No. 3 measures 500 ^ in several places.
The lack of special cytological preparations makes it difficult to homolo^i/.e the inier structure of the different tubule regions with those of other kidneys. In his description of the cell types in the two por- tions of tin- pniximal convolute of the sculpin, Grafflin (1937f) lists, among other>. the.M- three points of difference: the second part shows smaller nuclei, a more eo^'nophilic cytoplasm, and lower brush border, than the lir>t. Thi\ i^ exactly the reverse of the condition as regards these three characters in the Cyclothonc proximal convolute divisions. This suggests that the two portions of the proximal convolute in these two animals should be homologized in reverse order. Yet throughout the whole proximal convolute of the Cyclothonc nephron. the cells show
TELEOST KIDNEY IN CYCLOTHONE 361
a greater density of the cytoplasm on the luminal side of the nucleus, which is just the opposite of the condition reported in the sculpin.
In attempting to set up a general picture of homologies among the different fish kidneys, Grafflin (1937(/) concludes that the cell type of the second half of the proximal convolute of the sculpin is homologous with the same region in the tubule of the eel, and with the entire proximal convolute of the toadfish which shows no signs of division into two parts. Also, the first portion of the proximal convolute in eel and sculpin presents a second homology. Returning to Cyclothonc, if we try to make a comparison between the first part of the proximal convolute, and the second part of that segment in the eel, as we tried to do with the sculpin above, we find again the same similarity as to eosinophilic cytoplasm and lower brush border. The difference in nu- clear size, however, is not mentioned in Graffiin's description of the eel nephron (1937/?), and our comparison is further weakened by the appearance of his figure showing the point of transition between the two cell types. This photograph shows a much larger cell in the second portion of the tubule and suggests that the homology with Cyclothonc may well be a direct one, part one with part one and part two with part two. This view is further strengthened by the fact that a " zone of granular cells " sometimes appears in the first half of the proximal con- volute of the eel, just as does a similar zone in one of the Cyclothonc specimens (see above and Fig. 10).
The nephron of the lungfish, Lepidosiren parado.va, according to the description of Guyton (1935) is quite similar in size to that of the larger cyclothones, and except for having an initial collecting segment, it is divided into the same general regions. Grafflin finds that this tubule does not fit very well into his scheme of homologies among proximal convolutes. The first portion of that segment in the lungfish resembles the second in sculpin and eel, while the second part doesn't resemble anything at all in the other two. The principal cell in this second part has an inconspicuous brush border and lacks prominent vertical striations in the cytoplasm. While these characters make it as foreign to Cyclo- thonc as to the eel and sculpin, it is nevertheless worth noting that it shows a denser apical cytoplasm much like the cells of the proximal convolute in the Cyclothonc nephron. The shape and distribution (in- creasing toward the distal end) of the ciliated cells in the second part of that segment is similar in these two fish also.
It should be emphasized that the problem of homologies between cell types of the various segments in different tubules depends upon the constancy of the characters chosen for comparison. In the sculpin, the nuclei of the cells of the first part of the proximal convolute are
362 BRADFORD B. OWEN
described as being usually in a central position, although sometimes at the base of the cell. In the same region of the eel, nuclear position is stated as almost invariably basal. In the Cyclothone tubules, it is pos- sible to compare areas at exactly the same distance along each segment. \Yhen this is done one finds, using the first part of the proximal con- volute as an example-, a variation all the way from basal nuclei to apical ones (Figs. 7 and 8). The coarsely granular cytoplasm in the apical region of cells in both these tubules suggests that this area is engaged in active secretion, and that the nuclear position may simply vary with changes in tlu- cytoplasmic volume at the apical end. That such changes may follow a definite cycle can be seen from the histophysiological stud- ies made by von Mollendorff (1937) on fish kidneys. He observes a secivtory cycle involving changes in both cell border and cytoplasmic content. Such a series of repeated changes make it risky to consider many cytological features specific for one tubule region, until it be determined whether they are subject to change under various physio- logical states of the segment in question.
In conclusion I wish to state my sincere appreciation for the kind- ness of Professor A. B. Dawson of Harvard University in suggesting this problem and in offering valuable advice at all times in the course of the work.
SUMMARY
The Cyclothone kidney is extremely .simple. The entire system con- sists of two tubules running a practically .straight course side by side and uniting at the bladder. Thev are surrounded anteriorly by hemo- pi lie-tic tissue and posteriorly by the organ of Stannius.
The tubules can be divided into the following histologically distinct portions: glomerular capsule; neck segment; proximal convolute, with two types of brush border epithelium; ciliated intermediate segment; and terminal segment. An abrupt change in cell type separates the neck segment and the two parts of the proximal convolute, but the inter- mediate segment arises after a long zone of slow transition, and the cells of the terminal segment also appear gradually.
The cell types of the different regions of the nephron are described, with a discussion of the problems of homologies between tubule segments of different kidneys.
LITERATURE CITED
Auoicfe, J., 1910. Contribution a 1'ctude tics re-ins des poissons teleosteetis. Arch.
dc Zool. cxp. ct gen., 5'"" series, 4: _'75-o_'4. GRAFFLIN*, A. L., 1929. The jjpbudoglomeruli of the kidney of Lophius piscatorius.
Am. Join: Anal., 44: 441-454.
TELEOST KIDNEY IN CYCLOTHONE 363
GRAFFLIN, A. L., 1937«. The problem of adaptation to fresh and salt water in the teleosts, viewed from the standpoint of the structure of the renal tubules. Jour. Cell. Comp. Physio!., 9: 469-476.
GRAFFLIN, A. L., 1937^. Observations upon the structure of the nephron in the common eel. Am. Jour. Anat., 61: 21-62.
GRAFFLIN, A. L., 1937c. The structure of the nephron in the sculpin, Myoxo- cephalus octodecimspinosus. Anat. Rec., 68: 145-163.
GRAFFLIN, A. L., 1937</. The structure of the nephron in fishes. Representative types 'of nephron encountered; the problem of homologies among the dif- ferentiated portions of the proximal convoluted segment. Anat. Rec., 68: 287-303.
GUYTON, J. S., 1935. The structure of the nephron in the South American lung- fish, Lepidosiren paradoxa. Anat. Rec., 63: 213-229.
MARSHALL, E. K., JR., 1934. The comparative physiology of the kidney in relation to theories of renal secretion. Physiol. Rev., 14: 133-159.
MOLLENDORFF, W. VON, 1937. Zur Histophysiologie der Nieren von Hippocampus guttulatus und Lepadogaster candollii. Zeitschr. Zelljorsch. Mikrosk. Anat., 24: (1) : 204-226.
NASH, J., 1931. The number and size of glomeruli in the kidneys of fishes, with observations on the morphology of the renal tubules of fishes. Am. J. Anat., 47: 425-446.
PRICE, G. C., 1910. The structure and function of the adult head kidney of Bdello- stoma stouti. Jour. Ex per. Zoo!., 9: 852-864.
SMITH, H. W., 1932, Water regulation and its evolution in the fishes. Quart. Rev. Bio!., 7: 1-26.
VILTER, R. W., 1935. The morphology and development of the metanephric glomerulus' in the pigeon. Anat. Rec., 63: 371-385.
EYES OF DEEP-SEA CRUSTACEANS II. SERGESTIDAE
J. H. WELSH AND F. A. GRACE, JR.
(From the Woods Hole Occanographic Institution l and the Biological Laboratories, Harvard University)
INTRODUCTION
In the first of this series of papers which dealt with the eyes of the Acanthephyridae (Welsh and Chace. 1937) it was shown that Hyinawdora glacialis, a species of deep-sea prawn which normally lives at a level below the photic zone, has quite degenerate eyes, while related species which normally inhabit the photic zone have well-developed and obviously functional eyes. It was also shown that among the acan- thephyrids of the photic zone those possessing photophores have larger eyes in proportion to body size than those which lack the means to produce light of their own. Since such correlations between the struc- tural development of the eye, light intensity and the ability to produce light are of considerable interest from the standpoint of adaptation, the present study of the eyes of the Sergestidae was undertaken with these points in mind.
.Much of the literature dealing with the eyes of deep-sea crustaceans was cited in the paper already referred to and since collecting methods were also discussed in this paper and one by Welsh, Chace and Xunne- macher (1937), it will be unnecessary to discuss these matters further.
We are greatly indebted to Dr. H. B. Bigelow for the privilege of continuing thi> work and to the Milton Fund Committee of Harvard University for a -rant which cared for certain of the expenses of the investigation.
CHARACTERISTICS AND DISTRIBUTION OF THE SERGESTIDAE
The prawns which make up the family Sergestidae form a well- defined, aberrant group of the most primitive tribe of the Natantia, the Penaeidea. Since the species dealt with in this paper belong, in the main, to the genus Scryestcs it is not necessary to dwell on the four remaining genera of the group.
1 Contribution No. 176.
364
EYES OF DEEP-SEA CRUSTACEANS
365
All of the species of Scrgcstcs are slender-bodied animals with long, slender appendages (Fig. 1 ) and an integument that is never as firm as in most of the acanthephyrids. Despite their fragile appearance, most of the species retain their form and appendages fully as well as most other pelagic decapods when subjected to the usual handling. The last two pairs of thoracic legs are flattened and shorter than the other three pairs and fringed with long hairs. These, in conjunction with the long, well-developed pleopods suggest that the sergestids are among the fastest swimming of the pelagic Crustacea. Further indication that such is the case is provided by the fact that very few adult speci- mens and practically none of the mature individuals of the larger spe-
FIG. 1. Sergcstcs grandis Sund, female, showing certain of the features which characterize sergestids. Photophores, which are present only in certain species, are indicated by solid black spots.
cies were captured until the advent of the motor-driven vessel which was able to tow nets at a fairly rapid rate of speed. In collections now made with modern ships and apparatus, however, sergestids are very frequently encountered, and a number of the North Atlantic species are apparently reasonably abundant in the midwater area; possibly the commonest species in any one region is the luminous species, Sergestes lit ecus Hansen, from off Japan, of which 10 million pounds are said to be taken annually by commercial fisheries.
The lack of adult individuals and the extremely long and complicated larval history of these animals were the chief obstacles in the path of the correct determination of species by the early carcinologists. In- numerable species have been described, a great many on larval forms, and it was not until the last few years that any idea of the number of valid species could be learned. It is estimated that there are between
366 J. H. WELSH AND F. A. CHACE, JR.
thirty and thirty-five species of Scrgcstcs, and even now several com- parisons, particularly between Atlantic and Pacific specimens, are neces- sary before the exact number of known forms can be ascertained.
Although they inhabit the general " black fish-red prawn " area, only two of the sixteen species known from the North Atlantic are distinctly red in color. The others range from a deep rose pink and a pattern of red spots to almost perfectly transparent forms in which the internal organs may be clearly seen.
In 1903 photophores or luminescent organs were observed for the first time in any species of the genus and since then they have been described in four other species. Recently organs which are probably photophores have been detected in three other Atlantic species. It is very probable that further study of fresh material will show that about a dozen species possess these structures. So far as we are aware no instance of the discharge of a " luminous cloud " such as observed in several other bathypelagic forms has as yet been recorded for these forms.
The species are most abundant in tropical and sub-tropical areas, being most numerous in the Atlantic between the Tropic of Capricorn and 30° North Latitude. Although fourteen of the sixteen species known from the North Atlantic area were taken by " Atlantis " at Sta- tions 2666 (39° N., 70° W.) and 2667 (35° N., 69° 36' W.), only three species have been taken by any expeditions north of 47° N. Lat. The horizontal distribution of the species of Sergestes is similar to that of other bathypelagic decapods; several species have been taken at numerous localities throughout the world, while others have a restricted distribution with closely related species inhabiting corresponding areas in other oceans.
Comparatively little is known of the vertical distribution of these animals. This is far from surprising when one realizes that previous expeditions have for the most part used open nets, which arc perhaps more likely to catch animals when the net is being raised to the surface than at any other time. Also, most expeditions have covered a wide area and an attempt has been made in the reports to combine into one table the depth records for each species from these widely scattered localities. When one considers that probably no condition save pressure1 is everywhere constant at a given depth ; when one realizes that light intensity, temperature, oxygen content and the chemical composition of the water all vary markedly at different localities at a stated depth, it is no less than amazing that certain species are apparently found at so nearly similar depths in all parts of the area in which they are encoun- tered. Until exhaustive collections are made with closing nets at single
EYES OF DEEP-SEA CRUSTACEANS
367
stations and the results examined separately we will continue to be quite ignorant of the optimum conditions preferred by the various spe- cies. Table I gives a very rough idea of the depths at which the various
TABLE I
Depths in meters at which adult specimens of Sergesles have been taken with
closing nets by "Atlantis." 2
|
0 |
200 |
400 |
600 |
800 |
1000 |
1200 |
1400 |
1600 |
1800 |
2000 |
2050 |
|
|
Sergestes mollis3 |
X |
x |
x |
X |
x |
x |
x |
|||||
|
sargassi |
X |
x |
x |
|||||||||
|
pectinatus |
X |
X |
x |
x |
||||||||
|
arcticus . . . |
X |
x |
x |
x |
x |
|||||||
|
corniculum |
X |
x |
||||||||||
|
edwardsi |
x |
|||||||||||
|
cornutus |
X |
X |
||||||||||
|
vigilax atlanticus tenuiremis |
X X |
X X |
X X x |
X |
||||||||
|
grandis |
x |
x |
||||||||||
|
crassus |
X |
X |
||||||||||
|
robustus |
X |
X |
x |
x |
||||||||
2 The records of surface catches of adults of 5. atlanticus, S. cornutus and 5. vigilax have been taken from the reports of other expeditions.
3 The catches of 5. mollis at 600, 800 and 1,000 meters were all made at Station 2894. The lower transparency of the water in this region as compared with the Sargasso Sea, where most of the other hauls were made, may account, in part, for this occurrence of 5. mollis at depths so much shallower than those at which it is usually taken.
species may be encountered in the area explored by " Atlantis." The records for 5". arcticus were made at Stations 2463 and 2894, and for S. mollis at Stations 2475, 2666, 2667 and 2894; the other " Atlantis " specimens were all taken at Station 2667. Since most of the tows at Station 2667, from which most of the material was obtained, were made only at 400 and 800 meters, very little is known of the lower limits of any of these species. It is hoped that work on " Atlantis " in the near future at Station 2667 in greater depths will make the picture more nearly complete.
EXTERNAL FEATURES OF THE EYES
The accompanying figures (2-9) and Table II illustrate the amount of variation in the eyes of Sergestes. Since the body is less compressed and the bases of the antennal scales are broader in these forms than in
J. H. WELSH AND F. A. CHACE, JR.
the acanthephyrids, the eyes are borne on relatively longer stalks to enable the animal to see in all directions when the eyes are extended laterally. Since the length of the carapace is apparently not a con- stant fraction of the total body length and since such robust species as S. crassus have a much shorter carapace in proportionate body size than slender species like 6". corniculuin, it was found that the ratio of the diameter of the cornea to the carapace length showed less clearly the actual comparative sizes of the eyes than did the ratio of the diameter of the cornea to the entire length of the cornea and cyestalk and the
TABLE II
Relative size of the cornea as compared with the length of the carapace and the combined length of eyestalk and cornea in thirteen species of Sergestes.
|
Species |
Ratio Diameter of cornea to Length of cornea plus eyestalk |
Ratio Length of cornea to Length of cornea plus eyestalk |
Ratio Diameter of cornea to Length of carapace |
|
Sergestes mollis |
0 34 |
0.25 |
0.052 |
|
sargassi .... |
0 40 |
0.31 |
0.085 |
|
pectinatus .... |
045 |
0.36 |
0.099 |
|
arcticus . . |
0 46 |
0.38 |
0.090 |
|
corniculum |
n 17 |
0.43 |
0.072 |
|
edwardsi |
0 54 |
0.37 |
0.098 |
|
cornutus |
0 56 |
0.34 |
0.092 |
|
vigilax |
057 |
0.44 |
0.097 |
|
atlanticus |
0 58 |
0.42 |
0.105 |
|
tenuireniis* |
063 |
0.53 |
0.088 |
|
grandis*. |
063 |
0.57 |
0.081 |
|
crassus* |
0 65 |
0.50 |
0.132 |
|
robust us* |
067 |
0.67 |
0.115 |
* Luminescent species.
portion of the entire cornea and stalk occupied by the cornea. Even the latter comparisons are far from ideal as some species such as S. arcticus apparently have a longer stalk than others, but a more trust- worthy comparison could not be found. Every part of a prawn which might be compared to the diameter of the eye is found to show some specific variation.
It is apparent from the accompanying figures that by far the smallest eyes in proportion to body size are to be found in .V. mollis (Fig. 2), the only Atlantic species that obviously is restricted, for the most part, below the 1,000 meter zone. This species is a soft-bodied form, red in color, and it recalls to mind Hymcnodora among the acanthephyrids,
EYES OF DEEP-SEA CRUSTACEANS
369
although the eyes are far less degenerate than in the latter. One other sergestid, Pctalidium obcsnni, very probably inhabits a similar depth and the eyes in it are possibly even smaller than in S. mollis. The form
.
FIG. 2. Scrgestcs mollis Smith, male, 48 mm. in length. A species which is usually taken below the photic zone. Photophores absent. X 5.
FIG. 3. Scrycstcs arcticus Kroyer, male, 40 mm. in length. Photophores absent. X 5.
FIG. 4. Sergcstcs grandis Sund, female, 56 mm. in length. Photophores present. X 5.
FIG. 5. Sergestcs robustns Smith, male, 58 mm. in length. Photophores present. X 5.
with the largest eyes, 5". robustus (Fig. 5) is not, as might be expected, one of the species that frequents the surface layers. In fact, those four species which have the broadest eyes in proportion to the stalk, S.
370
J. H. WELSH AND F. A. CHACE, JR.
tcmiircmis (Fig. 6), 5". grandis (Fig. 4), ^'. crassiis (Fig. 8), and .9. robustus, are seldom taken in the upper 400 meters. Xext in order above these (possibly by chance) are the three species which have been
6
Fi<;. 6. Scrycstcs tcmiircmis Kroyer, male, 52 mm. in length. Pholophorcs
present. X 5.
FIG. 7. Scrycstes cornutus Kruyer, female, 14 mm. in length. Photophores
absent. X 5.
FIG. 8. Scrycstcs crassiis Hansen, male, 24 mm. in length. Photophores
present. X 5.
FIG. 9. Scrycstcs atlanticiis H. Milne Edwards, female, IS nun. in length. Photophores absent. X 5.
found at tin- --ill- face1, and between this group and .V. inollis are five species with rather small eyes which, so far as is known, frequent the lower part of the photic zone.4
•'The investigations >,\ \\YIsli, Chacc and Nunncmacher (1937), and more recent unpublished studies, indicate that certain of the sergestids undergo ex- tensive diurnal vertical migrations. The level of maximum numbers of a given species may be between the surface and 200 meters during the night and between 600 and 800 meters during the day. Therefore it is impossible to state that a certain species lives normally at a certain depth, and the most one can say is that the majority of the sergestids live in the photic zone and perhaps only one species in the North Atlantic, S. inollis, normally occurs below the level to which light penetrates.
EYES OF DEEP-SEA CRUSTACEANS 371
If these figures present a true picture of the relative sizes of the eyes iu the various species examined, one would naturally look for some factor to account for the large eyes of the four species at the bottom of the table. We believe that the answer may be found in the possi- bility that those species possess luminescent organs or photophores.
Until the sergestids of the " Michael Sars " expedition were re- ported upon by Sund in 1920 there was but one North Atlantic species, 6\ splcndcns Hansen, which was known to have photophores. Sund mentioned some pigmented spots in three other species, 5\ grandis, S. crassus (==S. splendens Sund), and 5". robustus, which he suggested might be luminous organs. From the material of these three species collected by ' Atlantis " we have been able to confirm his observations, and, although no histological preparations have as yet been made of this material, the position of the spots parallels so closely that of the photophores in those species which unquestionably possess them that there seems little doubt that such is their function. S. crassus and S. robustus apparently have these spots only on the antennal scales and outer uropods, while S. grandis (Fig. 1) has them scattered about the body and legs, particularly under the thorax and abdomen. In one large specimen of the latter species no less than 167 of these structures were counted. Hansen (1922) apparently doubted that these spots were actually photophores since they were not equipped with the lens-like structure found in the photophores of most other decapods and so become invisible in specimens preserved in alcohol or for too long in formalin. However, Kemp (1925) has described what he believes to be luminous organs in three species of pandalids from the Indian Ocean and these differ in structure from those which are known of other decapods in lacking any external trace of a lens.
If, then, the assumption is allowed that these structures are lumin- escent, only S. tcmdrcinis remains of those species which have the ratio of cornea to length of eyestalk greater than 0.60 and the ratio of the length of the cornea to the eyestalk greater than 0.50. Because no luminiscent structures were known in that species, a careful examina- tion of a large male specimen was made. Although the specimen had been in formalin for about seven months and the characteristic red color had consequently disappeared, a pair of large, whitish organs were found in the coxae of the last pair of thoracic legs near the open- ings of the vasa deferentia. Sections made from one of these organs proved that they were almost certainly a cluster of at least three large photophores with well developed lenses entirely enveloped by the sur- rounding tissues. Their position and the absence of an external lens
J. H. WELSH AND F. A. CHACE, JR.
closely parallels the structures described by Kemp in the Indian pandalids.
The evidence presented would indicate that there is some correlation between the size of the cornea in the species of Sergestes which live near the limit of light penetration and the presence or absence of luminescent organs. In the other five species which are known to have photophores, S. challcngeri Hansen, 5\ fulgcns Hansen, S. prchcnsilis Bate and S. splendcns Hansen the eyes are described and figured as large, although the figures are usually not sufficiently accurate to permit actual measure- ments. O. Pesta (1918) described structures on the inner side of the carapace in S. corniculum which he suggested might be photophores, but if these spots do prove to be luminous the proposed theory obviously does not hold for this species.5
Approximately one half of the described North Atlantic species possess a small tubercle on the inner margin of the eyestalk near the cornea. It may be of interest to list here the species which have or do not have this tubercle, although there seems to be no correlation be- tween the presence or absence of such a structure and the presence or absence of photophores or the size of the cornea.
Tubercle present Tubercle absent
5. armatus S. arcticus
S. corniculum S. atlanticus
S. crassus S. cornutus
S. grandis S. edwardsi
S. pectinatus S. mollis
S. sargassi S. robustus
S. tenuiremis S. splendens
S. vigilax
0 According to Burkenroad (1937) Organs of Pesta are probably present in all species of sergestids excepting S. mollis, S\ tenuiremis and S. challenge ri.
Figures 10-15 inclusive are all from dorso-ventral sections of eyes, unstained, X 32.
FIG. 10. Photomicrograph of a section of an eye of Sergestes corniculum made by means of transmitted light.
FIG. 11. The same preparation as in Fig. '10 photographed by means of dark- field illumination. The reflecting pigment which is more abundant in the dorsal part of the eye may be readily distinguished from the screening pigment.
FIG. 12. Photomicrograph of a section of an eye of Sergestes robustus made by means of transmitted light.
FIG. 13. The same preparation as in Fig. 12 photographed by means of dark- field illumination. The reflecting pigment is more abundant in the dorsal region which is toward the left.
FIG. 14. Photomicrograph of a section of an eye of Sergestes mollis made by means of transmitted light.
FIG. 15. The same preparation as in Fig. 14 photographed by means of dark- field illumination. The reflecting pigment is more evenly distributed throughout ihe region of the eye where it normally occurs.
EYES OF DEEP-SEA CRUSTACEANS
373
10
5-
-.. ^-.
14
FIGS. 10-15
374 |. H. \YKLSII AND !•". A. CHACK. JR.
Structure and Pigmentation of the E\cs
The eyes of eight species of M-r-estids were sectioned and examined histologically. The structure of these eyes was found to be essentially like that of deep and shallow-water shrimps and prawns whose eyes have been frequently described. One feature common to all was found to be the large and clearlv defined rhabdomes. the receptor dements of the eye. The most striking differences were seen in the amounts and distribution of screening and reflecting pigments.
The eyes of S. grandis and .V. tciutircinis resemble tho>e of S. robustus (Figs. \2 and 13). It has already been pointed out that these three >pecie> possess photo])hores and this probably accounts for their large eves. This mav also be the reason for the large' amount of reflecting ])igment found in these eyes. Its distribution is such that the periphery of the eve mav be more sensitive to differences in light intensity than the central portion, and the larger amount of proximal pigment in the central portion of the eye may conceivably increase the visual acuity of the region by a partial isolation of the rhabdomes from one another.
.S". corniciilitni ( Figs. 10 and 11 ). .S'. arcticns, .V. annul us and I'cta- lidiuin ohcsmn prohablv do not possess photophores and certain ot them may inhabit a region of higher light intensity than the three men- tioned above. ( n these forms only the first has any reflecting pigment in the' eye, the others completely lacking this set of pigment. \. arcticus and /'. ohcsmn have a large amount of black, proximal, screening pigment.
Since .S'. inollis normally lives below the photic /.one. as does the aran;he]ihyrid 1 1 yuicnodora i/lacialis. one might expect to find that its eyes were e(|uallv degenerate but such is not the rase. They are the smallest, in relation to holy M'/C. of all the sergestids examined by us. but structurally the eve shows verv little' modification which may be correlated with life in a /one to which sunlight does not penetrate i Figs. 14 and 15). The even distribution of retlccting pigment may ("iiceivably be related to tlie conditions under which the animal lives, but there are fairlv larg.- amounts of distal and proximal screening pigments. In I f\'iin'ii»(l<>ra (\\Vlsh and ('hare. l'>37) it was .shown that screening pigments were completely lacking, the rhabdomes had disappeared and there wa> an abnormally large amount of reflecting pigment. I leiice the eye was considered to be quite degenerate and capable of doing no more than registering changes in light intensity.
I )ecapod crustaceans which are known with certainty to dwell ex- clusively in the vast intermediate region of the sea, between the lower limit of the photic /one and the bottom ('excepting the bottom fauna).
EYES OF DEEP-SEA CRUSTACEANS
are not numerous as regards species. Until more have been studied it will be impossible to determine the exact trend in the degeneration of the eye as a result of living in complete darkness. From the work thus far carried out, however, it appears that depth, hence light intensity, modifies the eye, but in addition the possession of photophores and the ability to produce light is a most important factor in this adaptation.
SUMMARY
1. Fourteen species of Sergestes have been taken in closing nets from the western part of the North Atlantic and the size and struc- ture of their eyes have been related to the depth at which certain species occur and to the presence or absence of photophores.
2. Sergestes inollis is ordinarily taken below the photic zone and this species lacks photophores. The corneal portion of the eye of this form is smaller in relation to body size than is that of any other species studied. The eyes, however, are not so degenerate structurally as those of Hymenodora glacialis, an acanthephyrid having a similar vertical distribution.
3. Sergestes tenuiremis, S. grandis, S. crassus and S. robustus have been shown to possess organs which are probably photophores and these four species have the largest eyes in respect to body size of all which have been studied. Hence it may be concluded that the production of light by such an organism influences in some way the development of the eye. This agrees with the findings on the acanthephyrids.
4. The remainder of the sergestids studied which live within the photic zone have eyes smaller than those which possess photophores and in certain cases (S. arcticus and v$\ annatus) the pigmentation of the eye is quite unusual.
LITERATURE CITED
BURKENROAD, M. D., 1937. The Templeton Crocker Expedition. XII. Sergestidae (Crustacea Decapoda) from the Lower Calif ornian Region, with de- scriptions of two new species and some remarks on the Organs of Pesta in Sergestes. Zoologica, Sci. Contr. N. Y. Zool. Soc., 22: 315.
HANSEN, H. J., 1922. Crustaces decapodes (Sergestides) provenant des cam- pagnes des yachts Hirondelle et Princesse-Alice (1885-1915). Res. camp, sci. Monaco, 64: 1.
KEMP, S., 1925. Notes on Crustacea Decapoda in the Indian Museum. XVII. On various Caridea. Rec. hid. Mus. Calcutta, 27: 249.
PESTA, O., 1918. Die Decapodenfauna der Adria. Leipzig und Wien.
SUND, O., 1920. Penseides and Stenopides. Rcpt. Michael Sars N. Atlantic Deep- Sea Exf., 1910, 3: pt. 2 Zool.
WELSH, J. H. AND F. A. CHACE, JR., 1937. Eyes of deep-sea crustaceans. I. Acanthephyridae. Biol. Bull., 72: 57.
WELSH, J. H., F. A. CHACE, JR., AND R. F. NUNNEMACHER, 1937. The diurnal migration of deep-water animals. Biol. Bull., 73: 185.
THE QUANTITATIVE VERTICAL DISTRIBUTION OF
MACROZOOPLANKTON IN THE ATLANTIC
OCEAN BASIN1
BENJAMIN B. LEAVITT (From the Woods Hole Oceano graphic Institution, Woods Hole, Massachusetts)
During the summer of 1933 investigations were begun on the quan- titative distribution of plankton with depth in deep water in the offing of Woods Hole, based on the volumes caught at different depths at levels two hundred meters apart (Leavitt, 1935). The present report is a continuation of this work. It also includes a partial analysis of the fauna which is responsible for variations in quantity at different depths.
LOCALITIES OF COLLECTION
The collections were made from the research vessel " Atlantis," at 9 stations, 100 to 300 miles south and east of Woods Hole (Fig. 1) ; precise localities and dates are given in Table I (p. 378).
Three of the ten stations (1733, 2263, 2462) were located in Sargasso Sea water, offshore from the Gulf Stream, one (1735) in the axis of the latter,2 and six (1737, 1739, 2216, 2260, 2463, 2475) in the slope water inshore from the latter. These locations were chosen as representing oceanic conditions, all being well outside the continental edge, in deep water. One hundred and seventeen deep hauls were made, which may be dealt with from a quantitative point of view.
METHODS
The material was collected in closing-nets, either 1 meter or 2 meters in diameter, operated as earlier described (Leavitt, 1935). The releasing device (Fig. 2) has subsequently been improved to overcome the difficulties which arose with earlier models. It is made of phosphor- bronze and is hinged so that it may be opened to place it (on the cable) around a clamp which is bolted to the cable to prevent the releasing device from sliding up and clown. The toggle bars to which the bridles
1 Contribution No. 174 of the Woods Hole Occanographic Institution.
-Following Iselin (1933) the term "Gulf Stream" is used to include not only the band of warmest surface water or that wherein the northeastward current is most pronounced, but the whole body of water, extending down to 2,000 meters, across which the isotherms and isohalines dip steeply in the inshore-offshore direction.
376
\ KRT1CAL DISTUIIJUTION M ACROZOOPLANKTON
377
arc attached arc so shaped as to obviate any danger of their failing to swing out and drop the bridles, when released. The messengers, also, are hinged to facilitate placing them on the cable, and locked in place by a pin. The proper functioning of the net depends on the release of the toggle bars in the correct sequence by the messengers (Fig. 2). This depends not only on the second trigger being offset far enough to obviate any danger of its being struck by the first messenger, but also on the second messenger (cupping the first) being large enough in
70°
2475
CAPE COO
.1737
•2260
.1739
.2216
2463.
1735.
•CAPE HATTERAS
2462
'2263
BERMUDA
40
35
7 0°
FIG. 1. Location of stations.
diameter and long enough to receive proper guidance by the towing cable. Experience has also shown that unless the messengers are bored somewhat oversize the vibration of the cable may impede their progress downward through the water.
The releasing device and messengers are adapted for use in series and four nets were used, simultaneously, on the one cable.
Since there is no way of determining, from the condition of the nets themselves, on their arrival at the surface, whether they were open the full period of the tow, but only whether the proper bridles were
378
BENJAMIN B. LEAVITT
released, we must turn to the consistency of the results for evidence in this respect. It will be seen in Table1 1 that we usually found more than one rich level in the upper part of the water column, followed by two or more poor catches in mid-depths, with two or three somewhat more productive hauls still deeper. It is very unlikely that nets would consistently fish well at the shoaler levels, fail at the mid-levels, but work well still deeper down. Nor is it conceivable' that hauls from the same depths would fail every time at the same depths at different stations in different years. Stations 2216. 2260, and 2462 are particu- larly significant in this respect, for they show a marked similarity in vertical distribution (Table I). These similarities may be considered
TABLE I *
Catches in cubic centimeters, calculated per 2 hours towing with a 2-meter net
|
Sta. . |
1735 |
1737 |
1739 |
2216 |
IF 2260 |
2263 |
2462 |
2463 |
2475 |
|
|
Date. |
7/28/33 |
7/30/33 |
8/12/33 |
f. 1 1/34 |
8/23/34 |
7/27/35 |
7/31/35 |
8/31/35 |
Av- |
|
|
Lat.. |
36°50'N |
39°29'N |
39°42'N |
38°26'\ |
<•) 1 !'N |
35°17'N |
35°50'N |
37°21'N |
38°46' N |
er- |
|
Long. |
69°16'W |
70°14'W |
67°04'W |
69°05'W |
69°13'W |
67°2S'W |
68°39'W |
70°2(.'W |
7 1°38' W |
age |
|
200.\ |
105 |
300 |
250 |
.'0(1 |
93 |
33 |
105 |
375 |
v'l |
221 |
|
400. |
50 |
280 |
255 |
18 |
105 |
84 |
33 |
200 |
122 |
105 |
|
600. |
55 |
280 |
345 |
70 |
210 |
23 |
48 |
134 |
181 |
150 |
|
800. |
50 |
225 |
270 |
500 |
510 |
78 |
64 |
57 |
115 |
207 |
|
1000. |
33 |
125 |
150 |
45 |
10 |
75 |
42 |
54 |
148 |
77 |
|
1200. |
16 |
20 |
23 |
20 |
5 |
75 |
18 |
51 |
140 |
41 |
|
1400. |
37 |
57 |
27 |
55 |
90 |
35 |
4 |
35 |
8 |
39 |
|
1600 |
60 |
93 |
30 |
300 |
25 |
25 |
35 |
24 |
29 |
69 |
|
1800. |
80 |
125 |
45 |
70 |
3 |
8 |
10 |
9 |
0 |
31 |
|
2000. |
100 |
160 |
60 |
0 |
40 |
15 |
20 |
11 |
0 |
45 |
|
2200. |
105 |
10 |
60 |
38 |
17 |
5 |
||||
|
2400. |
151 |
12 |
32 |
62 |
4 |
|||||
|
2600. |
185 |
4 |
(, |
7 |
||||||
|
2800. |
220 |
2 |
4 |
4 |
||||||
|
3000. |
256 |
13 |
0 |
2 |
||||||
|
3200. |
0 |
2 |
* Positions and dates given are for the starting points of to\\ s.
strong evidence tor the adequacy of our technique since the first two stations were made two months apart, and the third a year later, all in the same general water mass.
Another proof of the accuracy of the method is found in the fact that in a great majority of the cases where hauls were repeated at the same level at the same station, the variation in the amount of material caught was !(•-,> than 30 per cent and often less than 10 per cent. In short, it seems that the methods employed were sufficiently reliable to yield consistent results.
The stated depths are as calculated from the length of win- out- board and its angle with no allowance for catenary. Failure to take account ot the latter no doubt introduces an error, the magnitude of which is assumed to be not greater than ± 100 meters. In any case,
VERTICAL DISTRIBUTION MACROZOOPLANKTON
379
it should he approximately the same for all hauls at a given level as care was taken to maintain a constant wire angle and control of the speed of the ship. The average speed of towing was close to two mile-, per hour in all cases.
The problem at hand being the difference in the productivity of the different levels of the sea, a volumetric system of analysis seems as significant as any other yet devised.
FIG. 2. Releasing device and messengers.
The total volumes of all the catches were first measured by the displacement method (Leavitt, 1935, p. 120), the volumes of the fol- lowing groups then measured separately : fish, copepods, decapods, euphausikls, chaetognaths, coelenterates, salps, and " residue."
Since the duration of the tows at different stations varied between one and two hours and since a few of the early tows were made with nets of different sizes, all volumes have been reduced to the common standard of two hours towing with a 2-meter net.
In addition to the volumetric measurements, the euphausikls were
380
BENJAMIN" B. LEAVITT
all identified and counted. In sonic cases, when great numbers were caught, it was necessary to use a sampling method for counting closely allied species (Leavitt, 1935, p. 125). Some other groups3 which were of particular interest in certain of the hauls have also been identified
400
600
800
1000
1200
1400
1600
1800
2000
2200
20 40
20 40 60
20 4 0
a
40 60 80
2400 2600 2800 3000
IK;. 3. Percentages of catch, at different levels at Station 221(>. .•/. Per- centage of total catch. B. Percentage of fish (•), copcpods (A), and cuphausiids (X). C. Percentage of decapods (O), eh;etognaths (A), and salps (X). D. Pcrcenta.u< of ccelentcrates (•), and "residue" (X).
and counted, cither in aliquot samples or in the entire catch, when the latter was small.
RESULTS
The most obvious feature of vertical distribution, illustrated by the foregoing tabulation, is that in spite of wide differences in detail
3 Drs. M. Sears, If. P.. Bigclow, and !•". Chacc have assisted with the identi- fication of copepods, coelenterates, and decapods.
VERTICAL DISTRIBUTION MACROZOOPLANKTON 381
from station to station, all stations agree in showing a peak of abundance at some level shoaler than 1,000 meters. And the validity of this generalization is corroborated by the fact that, on the whole, the per- centual distribution of the catches of the different groups at a given station, parallels that for the total catch there, as illustrated by Fig. 3 (derived from Table I).
In some instances, it is true, the curves for single groups show seemingly large peaks at depths where none appear on the curves for the total volumes. This is accounted for by the fact that the percentage of a single group at one level may be great when the importance of
TABLE II
Percentage taken
above 700 meters Depth at deepest haul
Station No. per cent meters
1735 58 2,000
1737 93 2,000
Percentage taken
above 800 meters
percent
1739* -- 3,200
2216 60 3,000
2260 80 2,200
2263 40 3,200
2462 55 2,400
2463 68 3,000
2475 85 2,000
* At Station 1739, the only station at which there were a great many salpae, over 94 per cent of the plankton caught was taken between 800 meters and the surface. If we subtract the volume of salpte which constituted over 80 per cent of the total catch in four out of the thirteen hauls, approximately 74 per cent of the catch was taken between 800 meters and the surface, 26 per cent in depths greater than 800 meters.
that group in the total community is negligible. Such instances are, however, exceptional.
The prevailing richness of the upper 800 meters, contrasted with the barrenness of the underlying stratum, is further emphasized by the following tabulation of the percentage of the total catch that was taken in the upper part of the water column (Table II).
It seems sufficiently established from Table II that from 40 per cent to over 90 per cent of the animals (by volume) were living in depths less than 800 meters, which is not surprising in view of the works of many early investigators. Within this upper zone the curves for individual stations differ widely, one from another, in the presence or absence of pronounced peaks (of abundance), and in the depths at
382 BENJAMIN B. LEAVITT
\vhich such peaks occur (Stations 2216, 2260, 2463, 2475, and to a lesser extent at 1739). Such peaks, in the upper levels, \vhich include the photic zone, may reflect the recognized vertical migration of some surface animals, the gregarious habits of others, as well as the fact that the phytoplankton is confined to this layer, so that one would therefore expect a more erratic vertical distribution of the quantity of plankton there.
Strongly contrasting with the abundant peopling above, is the fact that at every station, the water contained much smaller volumes of plankton either at 1,200 meters, at 1,400 meters, or at 1,800 meters, than at any shoaler level, appearing on the individual curves as distinct minima. However, the comparatively barren zone still contained a small amount of plankton, composed largely of dead and decomposing corpso in many instances, but in several cases (particularly Stations 1735, 2216, 2260, and 2462), there were living copepods. chaetognaths, decapods, and an occasional amphipod and aelenterate present, though in very small numbers.
Perhaps the most significant results of our exploration is that in 8 out of 9 cases, at least one level still deeper down proved considerably richer again, in plankton, either as a distinct maximum (Station 2216), as two such maxima (Stations 2260, 2465), or as a progressive increase in abundance, extending downward through a stratum several hundred meters in thickness (Stations 1735, 1737, 1739). In the extreme case (Station 2462), the secondary maximum at 2,400 meters was in fact some sixteen times as rich as the overlying minumum ; in another case (Station 2216), the tow was fifteen times as productive at 1,600 meters as at 1,200 meters.
Corresponding to this, the average catch, for all stations combined, \va> about 1.5 to 1.8 times as large at 1,600 meters as at 1,200-1,400 meters. And while averages for greater depths perhaps are not sig- nificant, because only four of the stations were worked deeper than 2. -Jin i meters, calculation of what percentage of the catch for the entire column was made at different depths shows a similar relationship (Fig. -li, followed by decrease from 1,600 meters to 2,000. Deeper than this, the average percentages were consistently insignificant down to the greatest depths fished, except at one station (1739) where the stratum between 2.-|no and 3,000 meters rivalled the upper 800 meters in richness. Even in that case, however, the deepest haul (3,200 me- ters) yielded nothing at all, although the net appeared to have fished properly. At two other stations (2263, 2463), also depths greater than 2,400 meters were practically barren, and conditions at Station
VERTICAL DISTRIBUTION MACROZOOPLANKTON
383
2475 show that, on occasion, this may apply at as small a depth as 1,800-2.000 meters (for details, see Table I. p. 378).
It is also significant — as hearing on its origin — that in every case the existence of a comparatively rich stratum at great depths, reflects,
Mo°-
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
2400
2600
2800
3000
10
15 20 25 %
\
FIG. 4. Percentages of total catch to different levels based on data in Table I. •, percentage at nine stations to 2,000 meters. X, percentage at four stations to 2,000 meters, v, percentage at four stations to 2,800 meters.
not only the relative volumetric abundance of the community as a whole, but also that of several individual groups of animals, as typified by the recorded volumes of fish, copepods, euphausiids, decapods, chaetog- naths, ccelenterates, salps, and "residue" at Station 2216 (Fig. 3). That these maxima represent widely diversified communities is further
384 BENJAMIN B. LEAVITT
proven by the large number of species of copepods and decapods repre- sented in hauls at the depths in question, at representative stations, as follows :
Station 1735, 2,000 meters, copepods, immature Mctridia sp., 75 per cent of sample, also Mctridia longa, Rhincalanus cornntus, Rhin- calanus nasutus, Paracuchata norvcgica, Paraeuchccta barbata.
Station 1737. 2.000 meters, immature Mctridia sp.. 24 per cent of sample, also the .-juries listed above, and Liiciciitia graudis, and decapods, Eucopia biunguiculata, Hymenodora sp.
Station 1739. 2,800-3,000 meters, copepods, immature Metridia sp., 37 per cent of -ample, also Calanns sp., Candacia sp., Corycceus sp., Eucalamts >]>., Euchccta sp., Euchirella sp., Gcctaiius sp.. Lucicittia sp., Oithona sp., Onccca sp., Phyllopus sp., Metridia luccns, Mctridia prin- ceps, Pleuromamma abdominalis, Pleuromamma gracilis, Pleuromamma robusta, Pleuromamma xiphias, Pscudochirclla obtusa, Rhincalanns cor- nutus, Rhincalanus nasutus, Scaphocalanus sp., Scolecithrix dame, Scot- tocalanus securifrons, Temora stylifcra, I'aldiviclla insignis.
Station 2216, 1,600 meters, copepods, immature Paraeuchccta sp., 51 per cent of sample, also, Euchirella intermedia, Euchirella messinen- sis, Euchirella rostrata, Hctcrostylitcs hngicornis, Mctridia princcps. Paraeuchtfta barbata, Paraeuchccta nori'ctjica, Pleuromamma robusta, Pleuromamma xiphias, Pscudochirclla notocantlia, Rhincalanus cor- nutus, Rhincalanus nasutus; decapods, Acanthcphyra purpurca, Ephy- rina bencdictcr, Eucopia sp., Gcnnadits sp., Hymenodora ghicialis, Scr- gestcs sp., and mysids.
Station 2462, 2,400 meters, copepods, Euchccta sp., 20 per cent of -ample, Lucicutia graudis, 11 per cent, Paracuchtcta barbata, 14 per i (lit, Rhincalanus nasutus, 17 per cent, also, Euchccta marina, Gaidius sp., Mcijacalanus sp., Metridia princeps, Paracuchata norvcgica, Pscu- dochirclla paliata, Pscudochirella pustUifera, Rhincalanus cornutus: decapods, 1 Gnathophausia sp., 1 Parapasipha sulcatifrons ; ca-lenterates. 1 I lalicrcas papillosum, 1 Homconcma sp., 6 Periphylla hyacinthina.
Station 2475, 1,600 meters, copepods. Rhincalanns nasutus, 30 per cent of total i-o])c]iods (310), Calanns sp. (immature), and Paraeuchccta norvcgica, together. 40 per cent; the remainder represented 15 other species, all of which have previously been found at similar depths.
Other haul- at the levels of the secondary maxima (Station 2260, 2,200 meters. Station 22(>1, 2,200 meters, Station 2462, 1,600 meters, Station 2463, 2,000 m< similarly yielded from one to seven speci-
mens each among the following oelenterates and ctenophores, JEyinura grinialdii. .-Imphicaryon acaiilc. .Holla bairdii. Kcrcc sp., Chuniphyes nniltidoilala, Ilalicrcas minimum, Hippopodius hippopus, Homconcma
VERTICAL DISTRIBUTION MACROZOOPLANKTON
sp., Pantachogon nihnini, J'cripliylla liyacintliina, and Rhopalonenia fnncrariuni.
VERTICAL DISTRIBUTION OF EUPHAUSIIDS
The euphausikls have been chosen as a representative group of animals for numerical analysis for the reason that they are holoplank- tonic and are worldwide in their distribution as a group, that they extend their vertical distribution to all depths of the ocean, that the species are not so numerous but what one may readily become familiar with them, and that little is known of their quantitative vertical dis- tribution outside the continental shelves.
Evidence that our hauls contained a representation of species suffi- ciently broad to be accepted as typical of mid-latitudes, on the high seas, is that they yielded all but one of the 27 species recorded by Tattersall (1926) from the collections made in 1914 by the U. S. C. G. Steamer " Bache " in waters adjacent to and south of the positions where the present collections were made. And this one (Thysanoessa microph- thalma) appears to be confined to lower latitudes of the Atlantic and Pacific, north and south (Hansen, 1915). On the other hand, our col- lection contains eleven which the " Bache " failed to find, namely, Thysanopoda oriental is, T. obtusifrons, T. acutifrons, Euphausia pseu- dogibba, Thysanoessa parva, T. inermis, T. longicaudata, Nematoscelis atlantica, Stylochciron affinc, S. insulare.4
Similarly, the present collections contain all the species reported by Hansen (1912) from the same general region, besides twelve others which he 'does not record there, although several of these are known to be widely distributed in the western Atlantic, and have been taken not far away; namely, Thysanopoda tricuspidata, T. pectinata, T. ob- tusifrons, Euphausia pscudoglbba, Ncmatobrachion fiexipes, N. boopis, Stylocheiron longiconic, S. abbreviatum, S. affine, S. maximum, S. suhniii, and S. insiihwc. Seven of these are recorded from farther south by Tattersall (1926). Again, we find, in the "Atlantis" col- lection, all but two of the species reported by Ruud (1936) from south- west of Ireland, from Cadiz Bay, and from the Mediterranean. And one of these (T. microphthalma) had already been reported from the western North Atlantic by Tattersall (1926), while the other (Nycti- phanes couchii} is a coastal species, so far known only in the Mediter- ranean and the eastern Atlantic.
We may conclude from the foregoing comparisons that we are dealing with a good representation of the total euphausikl fauna of the
4 Meganyctiphancs iwrvci/ica was also taken in several tows at stations 17. from unkown depths.
386
BENJAMIN B. LEAVITT
'~
c
' I
•c
N
CM
c"
V>
^ s J -a.
« -£
8
X
ft
c
•',
8
uo O
00
\O — O "*
00
OO
00
Ov
-t-f O
oo
-r
8
8
•
V
— CN
•— ' O OC
fOOO — ^-r^)^- o
C:fN
CN CN C
r-T
00
m
CN
OO
•^r^oO"*'-'OOO\ O CN •* <^>
-t
OO
••: •
' '
'z
3
2 :s ^ ^41 s 2 •
. - -^ •? 5? § t £= 5f.s:
j;i'SOcoocQ,2.2C5CC'C2!5OO<5<o1o
VERTICAL DISTRIBUTION MACROZOOPLANKTON
387
OO
(M
§
CN
CN
"53
•^ *-i
s
O O
w
J W <
H
SO
(N
10
C -H
OO
cu
JD
K
Q
OO •<* 00 >-H
ON ^-i -H
OOOs
•3 .5 g v g a ;
•?x,
C.
^- 1 \O
OO
2 c 9 ^
BENJAMIN B. LEAVITT
western Atlantic basin, especially in the fact that the list of species is a combination of northern with southern forms, to be explained by the latitudes at which the collections were made.
Earlier data (Leavitt. 1935. Tables II-V), and Tal.K- III of this paper show that the numbers of euphausiids taken was much smaller at great depths than at the shallower levels, for 86 per cent of the 76,790 specimens identified and counted were from tows between 800 meters and the surface, whereas only 14 per cent were from depths greater than 800 meters. These percentages are similar to those of the total volumes (p. 381).
Many of the euphausiids are notably gregarious animals and this fact coupled with vertical diurnal migrations may account for the unevenness of their occurrence in the upper layers.
In order to determine the relative importance of euphausiids as a whole in the community, the numbers of euphausiids per 50 cc. of plankton caught at each depth was estimated at four stations (Leavitt, 1935). While the numbers of euphausiids, like the quantities of plank- ton, increased on the average below 1,200 meters, the importance of this group in the total animal community decreased down to the deepest layer sampled, the number of euphausiids per 50 cc. of plankton being only about one-fourth as great at depths greater than 800 meters than in the upper layers.
There are few reliable data on the depth at which different species of euphausiids are to be found and Ruud (1936) regards temperature as one of the most important limiting factors of the distribution of the different species of this group. Since all but nine of the species recorded by him belong to groups most abundant between 200 meters and the surface, and since our shallowest hauls were at 200 meters, the data can hardly be compared with his. But among the nine species that In- ir-arded as deep-sea forms, it is interesting to note that T. (/rcf/ariu (n-garded by Zimmcr (1914) as bipolar) occurred abundantly at a depth of 2.000 meters at Station 1735, and T. parva, which did not occur in Kuud's material, may be regarded as even more persistently bathypelagic, on the basis of the present catches.
The divre.-ise in numbers of euphausiids with increasing depth is accompanied by a corresponding decrease in the specific diversity of the group, for Table 111 shows only 18 species out of the total of 37, as occurring at all, below 800 meters and most of them in such small numbers (sometimes a single' specimen) that only Bentheuphausia ant- hl\ops and Thysanocssa pan'a can be regarded as truly bathypelagic in the.se waters.
VERTICAL DISTRIBUTION MACROZOOPLANKTON
Thus, it seems sufficiently established, at least for this part of the ocean, and time of year, that euphausiids as a group reach their highest development in the upper 800 meters.
In the main, however, it is obvious, in common with Tattersall's (1926) experience, that the species in question show a deeper vertical range in the area investigated than in the eastern Atlantic and Medi- terranean areas surveyed by the " Thor ' (Ruud, 1936).
DISCUSSION Vertical D is t rib 1 1 1 io 1 i
The precipitous variation in the abundance of a species in a given area has often been remarked. Michael's (1911, 1913) and Hardy's (1936) statistics as to the variation in numbers of chaetognaths and other animals caught in a single place with successive hauls, and in closely proximate places, emphasize the fact that generalizations concerning distribution are extremely hazardous. Hence, we must realize the dynamic character of the plankton, and that attempts at a hard and fast terminology of the regions of vertical distribution are futile. It is only by continued observations over a period of time with a great many hauls that general ideas of distribution can be arrived at. Single hauls at separate localities can only be regarded as proof that what was caught was present at that particular time and place where the tow was made.
Nevertheless, it seems established by our investigations that in the region examined the abundant surface fauna decreases in quantity to a comparatively barren zone at a depth somewhere between 1,200 and 1,800 meters and that — in this particular region, at least — this minimum is followed by a subsequent increase in volume, at some greater depth.
Broadly considered, this is in line with previous knowledge, based largely on the work of Chun (1888, 1895), Alexander Agassiz (1888, 1892, 1902), Haeckel (1890), and Fowler (1898, 1904) that while pelagic animals live at all depths in the sea, with wide variety of species even at great depths, a much larger quantity occurs in the upper 800 meters or so than in deeper water.
Fowler's (1904) data on plankton from the Bay of Biscay is of particular interest in the present connection as the first detailed study of vertical distribution down to great depths in a restricted area, based on closing net hauls.
Subsequent investigators, including Esterly (1912), Murray and Hjort (1912), Michael (1913), Jespersen (1915, 1923, 1935), Hardy and Gunther (1936), Leloup and Hentschel (1935), Thiel (1935),
390 BENJAMIN B. LEAYITT
Ramner (1935), and Bigelow and Sears (1937), have also recorded vertical variations in quantity for siphonophores, medusae, copepods, fish, chaetognaths, corycaeens, and plankton as a whole. In none of these cases are the data of a sort to show whether or not there was any consistent increase in tin- volumes of plankton with increasing depth below the 1, 200-1, 800-meter level.
The plankton-poor layer encountered hy "Atlantis" somewhere between 1,200 and 1,800 meters is presaged by A. Agassiz (1888) and others. The rich streak at greater depths, however, was unexpected, and the animals found at these greater depths are holoplanktonic. not benthonic, as illustrated by the assemblage collected at Station 1739 from a depth of 3,200 meters, where the total depth was 3.674 meters. In fact, there was no indication of the presence of benthonic animals, in any of our hauls, nor, in fact, would such have been expected for even the deepest tows were still far above the bottom.
In view of the changes in the physical and chemical characteristics of water with increasing depths, it is interesting, in the first instance. to correlate the volumetric distribution, vertically, of the plankton of the open sea with the density, salinity, and temperature at different levels in the slope water and in the (iulf Stream beyond.
Typical graphs (Fig. 5) for temperature, salinity, and density for the waters in which these investigations were carried on show no correla- tion between any of these environmental factors, and the depth distribu- tion of the animals. Thus, the temperatures at all depths are well within the range in which we find plankton thriving at other times and places. The salinity variations are too minute for us to suppose that they have any appreciable effect on the plankton in tliese waters. Again, we find the same animals living throughout wider ranges of salinity than are pre-'iited here. The density does not show any indication of layer- ing such as might account for an accumulation of material at any par- ticular level between the surface and the bottom. In the present case, there is no apparent correlation between quantities of plankton, and either oxygen or light. Neither do any of these characters of the water show fluctuations below a depth of 800 meters corresponding to the existence of the plankton minimum at 1.200-1.800 meters, or to the subsequent increases in the volume of plankton at some greater depth.
The minimum oxygen in cubic centimeters per liter does not fall below 60 per cent saturation or about 3.5 cc. per liter in any of the water which was fished ( Seiwell. 1934). And comparison between data for oxygen and numbers of animals caught at many of the " Me- teor " stations as reported by \Vattenberg (1935), Spicss (1932), and Hcntschel (1932) shows that an abundance was caught in water where
YHRTICAI. DISTRIBUTION M \CK( )X( )()I'LANKT()N
391
the oxygen content was much lower than at any level in the waters with which the present investigations are concerned. We must, there- fore, conclude that there is ample oxygen to support planktonic commu- nities at all depths there.
10 12 14 16 16 20 22 24 26°C
340 \ 350/V
10 12 14 16 18 20 22 24 2 6 C
FIG. 5. Temperature, salinity, and density at " Atlantis " Station 1729, typical of slope water between the Continental Shelf and Gulf Stream and at Station 1734, typical of Sargasso Sea water east of the Gulf Stream, in the area of these in- vestigations. T, temperature in °C. S, salinity in parts per mille. D, density (specific gravity minus one, times one thousand).
Among other possible causes for such vertical variations in abun- dance of plankton as do occur in deep water, the following possibilities come to mind :
392 BENJAMIN I',. l.KAYITT
(a) Dead and decomposing animals sifting down from the surface arc sometimes caught in deep hauls. This happens oftenest with lasis zonaria. the firmest bodied salp, pieces of which are found among the other dead and decomposing animal matter in deep hauls. Other sal- pidae, also, contain cellulose which might be expected to disintegrate slowly.
(/>) An occasional large medusa, Periphylla hyacinthina or Atolla bairdii, or some large deep-sea decapod, may rarely be responsible for a relatively voluminous catch at a great depth.
(r) A wc-11 developed bathypelagic fauna may — for some undeter- mined reason — find some particular deep levels more suitable as an abode than the levels above or below. The general identity of the animals taken in our tows shows this not to be the case.
(d) Sinking water masses from more moderate depths, in somewhat higher latitudes may bring their planktonic communities with them and so be responsible for richer strata below poorer, farther south. This last seems the most probable explanation for such a distribution in the present case, because species of decapods, copepods, ccelenterates, and euphausiids occurring at moderate depths somewhat further north have been prominent in the catches where total volumes have revealed the existence of secondary maxima, deep down. This, for example, ap- plies to Calanns finmarchicus, Mctridia longa, Paraeuchccta norvegica, Rhincalanus nasutns, Rh'mcalamts connilus, Pleuromamma abdominalis. Pleuromamma gracilis, Pleuromamma robusta, Pleuromamma xiphias, among the copepods. several species of Gennadus, among the decapods, and Thysanoessa gregaria, among the euphausiids.
It is likely that the animals belonging in this category are able to maintain themselves even in the deepest water layers for a considerable length of time, and that great depths may constitute the outskirts of the areas which they regularly populate. Lack of light may not harm them. Dr. C. B. Wilson has informed me that his examination of co] )e])od material brought back from the Ryrd Antarctic Expedition showed that many (including larval forms) were taken from under the ice, where it was totally dark, and far from open water. The " Terra Nova" Kxpedition also obtained many thousands of specimens under the ice (Far ran. ]'>->). In short, communities may maintain themselves for a long time after sinking. We may have a biological loss, akin to that of the animals and plants that drift north and to their deaths in the surface waters of the < inlf Stream every summer.
While examination of vertical distribution was our primary objective, one phase of the horizontal picture deserves at least passing mention, namely the poverty of the catches in the Sargasso Sea water and in the
VERTICAL DISTRIBUTION MACROZOOPLANKTON
Gulf Stream, as defined on p. 376, contrasted with the slope water, the average catch (to 2,000 meters) being only 415 cc. in the first of these regions, 586 cc. in the second, contrasted with 1,283 cc. in the third. This regional gradient is of the same order as reported by Jespersen (1923, p. 8, Fig. 23) for average volumes for the upper 600 meters farther to the south, and is also in line with the prevailing poverty of the Sargasso Sea region, subsequently demonstrated by him (Jespersen, 1935, Figs. 27, 28).
BIBLIOGRAPHY
AGASSIZ, A., 1888. Three cruises of the "Blake." Bull. Mus. Compar. Zool., 14:
1; 15: 1. AGASSIZ, A., 1892. General sketch of the expedition of the "Albatross," from
February to May, 1891. Bull. Mus. Compar. Zool., 23 (No. 1) : 1. AGASSIZ, A., 1902. Preliminary report and list of stations ("Albatross" 1899 and
1900). Mem. Mus. Compar. Zool, 26: No. 1, p. 1. BIGELOW, H. B., AND MARY SEARS, 1937. Siphonopliorse. Kept. Danish Oceanog.
Expcd., 1908-1910 to the Mediterranean and Adjacent Seas. Vol. 2, Pt.
2, p. 1. CHUN, C, 1888. Die pelagische Thierwelt in grosseren Meerestiefen und ihre
Beziehungen zu der Oberflachenfauna. Bibliotheca Zoologica, 1 (Heft
1): 1. CHUN, C., 1895. Biologische Studien iiber pelagische Organismen. Bibliotheca
Zoologica. Heft 19, Lieferung 1, p. 1. ESTERLY, C. O., 1912. The occurrence and vertical distribution of the Copepoda
of the San Diego region. Univ. Calif. Publ, Zool., 9 (No. 6) : 253-340. FARRAN, G. P., 1929. British Antarctic ("Terra Nova") Expedition, 1910.
Natural History Repts. Zoology, Vol. VIII, Crustacea, No. 3, Part X,
Copepoda, pp. 203-306. FOWLER, G. H., 1898. Contributions to our knowledge of the plankton of the
Faeroe Channel. No. VI. Proc. Zool. Soc. London, pp. 567-585. FOWLER, G. H., 1904. Biscayan plankton collected during a cruise of H. M. S.
"Research," 1900. Trans. Linn. Soc., London, 10 (Pt. I) : 1. HAECKEL, ERNST, 1890. Plankton Studien. Jcnaische Zeitschrift, XXV, and
translated by G. W. Field as Appendix 6 to the Report of the Commis- sioner for 1889, 1891, U. S. Com. Fish and Fisheries, Washington, 1893,
pp. 565-641. HANSEN, H. J., 1912. The Schizopoda. Reports . . . U. S. Fish Com. . . .
"Albatross." Mem. Mns. Compar. Zool., 35: 175-296.
HANSEN, H. J., 1915. The Crustacea Euphausiacea of the United States Na- tional Museum. Proc. U. S. Nat. Mns., 48: 59-114. HARDY, A. C., 1936. Observations on the uneven distribution of oceanic plankton.
Discovery Reports, 11: 511-538. HARDY, A. C., AND E. R. GUNTHER, 1935. The plankton of the South Georgia
whaling grounds and adjacent waters, 1926-1927. Discovery Reports,
11: 1^56. HENTSCHEL, E., 1932. Die biologischen Methoden und das biologische Beobach-
tungsmaterial der " Meteor " Expedition. Wiss. Ergeb. Deutschen Atlant.
Exped. "Meteor," 1925-1927. Vols. 10 and 11. Berlin and Leipzig. ISELIN, C. O'D., 1933. The development of our conception of the Gulf Streams
system. Trans. Am. Geophys. Union Fourteenth Annual Meeting, pp.
226-231.
394 BENJAMIN B. LEAYITT
JESPERSEN, P., 1915. Sternoptychidae (Argyropelecus and Sternoptyx). Kept.
Danish Oceanog. Exped. 1908-1910 to the Mediterranean and Adjacent
Seas. No. 3, Vol. II (Biol.). A. 2.. pp. 1-41. JESPERSEN, P., 1923. On the quantity of 'macroplancton in the Mediterranean
and the Atlantic. Rcpt. Danish Occanog. Exped. 1908-1910, No. 7, Vol.
Ill, Misc. papers, Pt. 3, pp. 1-17.
JESPERSEN, P., 1935. Quantitive investigations on the distribution of macroplanc- ton in different oceanic regions. Dana Report, No. 7, p. 1. LEAVITT, B. B., 1935. A quantitative study of the vertical distribution of the larger
zooplankton in deep water. Biol. Bull., 68: 115-130. LELOUP, E., AND HENTSCHEL, E., 1935. Die Verbreitung der calycophoren
Siphonophoren im Sudatalantischen Ozean. U7iss. Ergcb. Deutschen
Atlant. Exped. " Meteor," 1925-1927, Vol. 12, Pt. 2, pp. 1-31. MICHAEL, E. L., 1911. Classification and vertical distribution of the Chaetognatha
of the San Diego region. Univ. of Calif. Publ. Zoo!., 8: 21-186. MICHAEL, E. L., 1913. Vertical distribution of the Chaetognatha of the San
Diego region in relation to the question of isolation vs. coincidence. Am.
Nat., 47: 17-49. MURRAY, J., AND J. HJORT, 1912. The Depths of the Ocean. MacMillan and Co.
Ltd., London. RAMNER, W., 1935. Die Cladoceren der " Meteor " Expedition. Wiss. Ergcb.
Deutschen Atlant. E.vped. Meteor, 1925-1927, Vol. 12, Pt. I, pp. 111-121. Ruun, J. T., 1036. Euphausiacea. Reft. Danish Occanog. E.vped. 1908-1910 to
the Mediterranean and Adjacent Seas, Vol. II, D. 6, pp. 1-86. SEIWELL, H. R., 1934. The distribution of oxygen in the western basin of the
North Atlantic. Papers in Physical Oceanography and Meteorology
published by Massachusetts Institute of Technology and W. H. O. /.,
Vol. Ill, No. 1, pp. 1-86. SPIESS, F., 1932. Das Forschungsschiff und seine Reise. Wiss. Ergcb. Deutschen
Atlant. Exped. "Meteor" 1925-1927, Vol. I, pp. 1-413. TATTERSALL, W. M., 1926. Crustaceans of the Orders Euphausiacea and Mysidacea
from the western Atlantic. Proc. U. S. Nat. Mus., Vol. 69, Art. 8, pp.
1-31. THIEL, M. E., 1935. Die Besiedlung des Siidatlantischen Ozcans mit Hydro-
medusen. Wiss. Ergcb. Deutschen Atlant. Exped. "Meteor," 1925-1927,
Vol. 12, Pt. 2, pp. 32-100. WATTENBERG, H., 1935. Das chemische Beobachtungsmaterial und seine Gewin-
nung. Wiss. Ergeb. Deutschen Atlant. E.vped. "Meteor," 1925-1927,
Vol. 8, pp. 9-121.
ZIMMER, CARL, 1914. Die Schizopoden der Deutschen Siidpolar-Exp. 1901-1903. •//. Siidpol. Exped., Bd. 15. Zool. Bd. 7, pp. 377-446.
A RECONSIDERATION OF THE EVIDENCE CONCERNING
A DORSO-VENTRAL PRE-ORGANIZATION OE
THE EGG OF CH.ETOPTERUS
T. H. MORGAN
(I'roiii the Marine Riohujical Laboratory, U'oods Hole, Mass, and the William
Ci. Kcrckhoff Laboratories of the Biological Sciences, California Institute
of Technology, Pasadena. California)
In a former paper (1937) the question of the presence of a dorso- ventral axis in the unfertilized egg of Chcetoptenis was discussed. The argument rested in part on the assumption that the large germinal vesicle is excentric as in many other eggs, but here both laterally as well as axially excentric. Since, in eggs that have come to rest in a single layer on the bottom of a dish of sea water, the pole lies, as a rule, not at the top of the egg but more or less at the side, and since the germinal vesicle appeared, then, to lie symmetrically with respect to the circum- ference of the egg as seen from above (as in Fig. A), and since the first cleavage passes through the pole and in most cases through the upper- most side of the egg, it seemed to follow that there is a dorso-ventral predetermined orientation in the unfertilized egg.
Later, when I came to examine sections of the egg, it could not be visibly demonstrated that the germinal vesicle lies nearer the pole, i.e., in an excentric position with respect to the polar axis. It became evi- dent then that the problem of the location of the dorso-ventral axis called for a closer examination. This examination I have made during last summer. The new evidence shows that the former argument is insuf- ficient to establish the conclusions there presented.
If, then, the germinal vesicle does not lie excentric in the polar axis or excentric to that axis (as in Fig. A), it remains to consider other possible relations.
If, for example, the nucleus lies in the center of the egg (as in Fig. B} then when the pole lies above (or below) the equator of the egg the nucleus will again appear as seen from above to be in the center of the egg, and it is not possible to distinguish on this evidence alone, between A and B. There is a third possibility, viz., that the germinal vesicle lies excentrically in the polar axis (Fig. C) but not excentric to that axis. If this were true, then the egg, seen from above, would, when it lies with its polar axis oblique to the vertical, show the germinal vesicle
395
396
T. H. MORGAN
slightly excentric to the circumference of the egg, but this does not fit in with the new or the old observations.
Should the pole of the egg come to lie near the uppermost part of the egg, as the egg settles clown on the bottom of the dish, as indicated by any one of the ring of dots in Fig. /), then, the first cleavage plane, that always passes through the pole and the center of the egg, will be an approximately vertical plane. If the pole lies somewhere between the
I
I
Pole
Pole
Pole
l-'n.. 1. ./. diagram of egg showing nucleus excentric on primary axis and excentric to that axis. The arrow indicates the expected view as seen from above. /?, diagram "f egg with nucleus in center. C, diagram of egg with nucleus ex- centric mi primary axis, but not excentric to it. I), egg with pole uppermost, as indiiatril I iy line circle of dots. E, same with pole sideways between top of egg and hori/.un. /•'. same with pole at or near horizon.
top of the egg and the equator, as in Fig. /:. the first cleavage plane may be more or le>s oblique to the vertical, but will appear to cut through the upper Mirfaee of the egg in the great majority of cases. But if the pole of th' it or near the equator, i.e.. it" the polar axis lies nearly
in the horizontal plane, as in Fig. /;, then the first cleavage plane should appear horizontally as well as vertically and also in all intermediate positions. On the basis of chance alone there should be as many vertical as horizontal first cleavages when the polar axis lies horizontally. Such
PRE-ORGANIZATION OF CH^ETOPTERUS EGG 397
eggs should furnish critical evidence concerning the point at issue, viz., whether, as I assumed , the eggs come to rest with the dorso-ventral axis vertical.
As stated above, sections of eggs, killed at once after removal to sea water, showed a nucleus irregular in outline lying at or near the center of the egg. If it is excentric, the excentricity is so slight as to be negligible, and certainly would not be observable in the living egg. Since these eggs had been taken out in sea water and killed " at once," it was possible that the irregular outline of the nucleus was due to the action of the sea water which might make its excentricity difficult to detect. In fact the nucleus breaks down to form the polar spindle very soon after removal of the egg to sea water.
A new set of eggs wras then preserved in the summer of 1937, without coming in contact with sea water, in four different killing fluids (Flem- ming, Bouin, picro sulphuric, boiling water) and sectioned and stained. In some cases the eggs became free in the preservative ; in other cases they remained in the tied-off parapodium. The preparations showed that the parapodial eggs lie several layers deep on the string or band from which they arise, and, owing to crowding, have many different shapes ; but even those in the outermost layers are very seldom perfectly spherical. In none of them is the nucleus obviously excentric. The nuclear wall conforms in shape somewhat to the shape of the egg, and is not often perfectly spherical. After different preserving fluids (and subsequent reagents prior to imbedding), the shape of the nucleus is to some extent affected, but is seldom perfectly spherical. It follows then that when removed to sea water the irregular outline of the nucleus is not due to the absorption of water by the egg (as the egg assumes a spherical shape), but traces back to the parapodial eggs.
An examination of the living eggs as seen from the side was also made. The eggs were placed in a drop of sea water on a slide (ringed with vaseline) across which a thread had been stretched. The prepara- tion was held in a vertical position and examined under the microscope. If the nucleus lies in an excentric position on the polar axis this should become visible in some, at least, of the eggs, but no excentricity was observed. It is true that the outline of the nucleus of the living egg is not sharply denned, but nevertheless were the nucleus excentric this should be apparent, which is not the case. Of course, in a confined space (but not compressed) the eggs might not have time to orient as they fall onto the thread. Nevertheless, some of the eggs should show the excentricity of the nucleus if it were present.
A more critical examination of the position of the first cleavage was then made. It was found, as expected, that in most of the eggs the
398 T. H. MORGAN"
first cleavage was approximately vertical. These eggs had settled to the bottom of the dish containing sea water and were not covered. In an unexpected number of eggs the lobe appeared later on the upper surface, proving that in some eggs the polar hemisphere was down. Special attention was paid to those eggs in which the clear area (con- taining the pole) lay near the horizon, so that more eggs of this kind were selected. Since the polar area is difficult to see if it lies exactly above or below but easily seen if it lies nearer the equator, a larger number of eggs with an oblique polar axis (nearly horizontal) would be those selected for examination. In a total of 111 eggs recorded there were 58 in which the first cleavage was vertical, i.e.. through the top of the egg; there were 13 in which it was oblique; and 40 in which it was horizontal or approximately so. In about half of the latter the smaller blastomere was above (24 cases), and in half the larger one was above. Adding together the oblique (13) and the horizontal cleavages (40) gives a total of 53 eggs. In other words, in this selected group of eggs about half of the cleavage planes were not vertical but oblique or near the horizontal plane.
For comparison with these figures another count of unsclcctcd eggs in the first cleavage stage was made, recording those in which the first cleavage was vertical and those in which it was horizontal. In 120 eggs all but six showed the plane vertical or nearly so; three of the six had the large cell on top and three the smaller cell.
Since, as explained above, vertical cleavages should predominate on chance alone, as they do, and since, as anticipated, some of the eggs in which the pole lies at the side should show oblique or horizontal cleav- age, it follows that the original assumption that a definite side of the egg lies above is not established, but is the expected outcome of the relation of the cleavage plane to the pole.
A check on the number of eggs in which the pole lies below the hori/oii is obtainable in two ways in addition to observation of the polar bodies, which is often difficult. After the second cleavage the configuration of the four blastomeres is such, owing to the broad cross- furrow at the antipolar hemisphere, that the two poles can be distin- guished from each other. The other method of locating tin- pole and the antipole is from the position of the yolk lobe at the time of the first cleavage, when it is either up or down. When it lies under the egg it is not otten seen. In the records there are cases where the lobe was above, hence the pole was down, showing that it may sometimes lie below the horizon, i.e., such eggs have not oriented to gravity, but the observations were not systematically carried out. hence the ratios of cases where the nole is up or down cannot be definitely stated.
PRE-ORGANIZATION OF CH^TOPTERUS EGG 399
The evidence discussed above does not, of course, disprove the view that before fertilization there is a dorso-ventral organization of the egg. Whether there is such an arrangement still remains an open question to be examined by other methods, one of which should be a study of the relation of the first cleavage plane to the point of entrance of the spermatozoon. In addition to an earlier examination of this relation by Morgan and Tyler (1930) fifty-three new cases have been studied and will be described in another paper.
It remains to be discussed whether the polar hemisphere is lighter than the antipolar, which seemed to be demonstrated by the centrifuging experiments described in my former paper. The argument for this assumption was based both on the position of the pole when the eggs are centrifuged before the polar bodies are formed, and also on the orientation of the egg in the machine after the first or second polar body had been extruded. The evidence showed that while most eggs orient with the polar region toward the center of rotation, the orientation is by no means exact. The higher degree of orientation on the centrifuge than of orientation to gravity in a dish of shallow water is not sur- prising since the former force is greater than the latter, and also since the eggs, falling through a longer space, have more time to orient before reaching the bottom of the centrifuge tube than have the eggs in a dish of water. However, unless the eggs are centrifuged over a gum-solution on which they lie in a single layer, this inference does not hold entirely, since if many eggs are centrifuged in sea water, and come to be packed at the bottom of the tube, a considerable amount of shifting of the eggs may take place as they become crowded together.
If, then, it be admitted that the majority of eggs turn on the cen- trifuge with the polar (or animal) hemisphere toward the center of rotation, still this will not explain why the pole of the egg is as a rule excentric to the stratification. One possible explanation suggests itself, namely, that there is a ring of material around the pole that is lighter than the pole itself. This may seem improbable and is not demonstrated by the evidence at hand. The following hypothesis may seem more plausible. When centrifuging begins, the eggs lie at random with respect to the axis of rotation. Some of them will have their polar hemisphere more or less in this axis, others will have it at the side, and others will have the polar hemisphere away from the center of rotation. Under the influence of the centrifugal force all of the eggs will begin to turn with the polar hemisphere toward the axis of rotation. Now if the redistribution of the lighter and heavier material within the egg begins while the egg as a whole is rotating, its rotation will cease as soon as the redistribution of the materials brings a region lighter than
400 T. H. MORGAN
the pole toward the center of rotation. Hence many of the eggs may have the oil cap to one side of the true pole. viz.. those that were not centered when the movement of the material within the eggs began to take place. This suggestion could possibly be tested by allowing the eggs to orient to gravity before centrifuging. More of them would then be expected to show the pole in the center of the oil field.
In connection with the problems discussed above the question re- mains as to whether or not the egg shifts its position after settling on the bottom of the dish. The fact that the egg undergoes a definite series of changes in shape might seem to make it possible that a partial rotation might take place. For example, when the first polar body is extruded the polar cap flattens and the egg becomes broader, i.e., biscuit- shaped. It then rounds up and remains spherical until about five min- utes before the cleavage is due. when it becomes pear-shaped with the small end of the pear at the pole. Again it becomes spherical and flattens over the polar area as indications of the cleavage begin to appear. At this time it elongates laterally, i.e., it becomes biscuit-shaped as the lobe appears exactly opposite the polar bodies. At first it can not be foretold which side will become the smaller blastomere and which the larger, but very soon the difference becomes apparent in side view and this continues until a large and a small cell arc' formed. The egg is much extended laterally at this time. The two blastomeres next come together, but the egg does not become quite spherical. During the first of these changes in form the- eggs do not turn over as shown by the constant position of the polar bodies, but as the two blastomeres come together, the first cleavage plane may appear to be more nearly vertical than during the division. Kvidently the jelly around the egg adheres sufficiently to the dish to keep the egg in place. Moreover, the egg cannot rotate within the surrounding membrane. In fact, the latter adheres so closely to the egg that it indents with the cleavage furrow. It consists of a transparent inner layer — which may be a gelatinous layer between the egg and the outer jelly. When the polar bodies are extruded they push the inner membrane before them. I before and dur- ing this time the- membrane is thrown into a series of corrugations over the polar region. Similar corrugations appear later over the region where the lobe is about to appear. The meaning of this folding of the membrane is not apparent, since it might have rather been supposed that the membrane would be stretching over these regions. Possibly surface movements of the protoplasm toward the region of activity may cause the wrinkling of the membrane.
THE RELATION BETWEEN ENTRANCE POINT OF THE
SPERMATOZOON AND BILATERALITY OF THE
EGG OF CHJKTOPTERUS
T. II. MORGAN AND ALBERT TYLER
(From the Marine ttioloyical Laboratory, Woods Hole and the William- G.
KerckJioff Laboratories of the Biological Sciences, California Institute
of Technology, Pasadena, California)
In a preceding paper (Cytologia, 1937) we examined the relation of the entrance point of the spermatozoon to the cleavage planes of Ch&topterus, Cuiningia, and Nereis. In Chcctoptcrus there was a co- incidence of 41 per cent between the entrance point and the first plane of cleavage. In Cuiningia the coincidence was 79 per cent. In Nereis it was 51 per cent. The failure to obtain 100 per cent coincidence may be due to errors of observation or to other factors being involved in determining the first cleavage plane.
We have made some additional observations during the past summer on Chcetopterus since its eggs gave the lowest percentage of coinci- dence, and since a larger number of observations seemed desirable. Out of 53 new observations there were 24 cases of fairly close agree- ment ; in 7 cases the divergence was less than 45 degrees, and in 22 cases it was more than 45 degrees. Approximate coincidence was 45 per cent, as compared with 41 per cent of the old data. This absence of exact agreement led to an examination of other possible relations. When the entrance point is plotted on the egg in the two-cell stage it is found in the great majority of cases that it lies on the side of the egg where the smaller blastomere (AB) is found. In the new data there were 29 cases on the AB side, and only 9 cases on the CD side. In the older data there were 47 cases on the AB side, and 21 on the CD. But since the CD side has twice the surface area of the AB side the difference is more significant.
In these data cases of exact coincidence between the entrance point and the first cleavage plane are, of course, omitted. When the entrance point is plotted on the four-cell stage of these same eggs it is found that, in the majority of cases, it lies on the B cell, namely in 20 cases on the B ; in 3 cases on the A ; in 3 on the C ; and in two on the D cell. In the new data it lies on the B cell in 15 cases, in 9 cases on the A, in 7 cases on the C, and 2 on the D cell. Cases of coincidence with the first cleavage plane give 11 with D opposite the sperm and 2 adjacent.
401
402 T. H. MORGAN* AXD ALBERT TYLER
Taking these results as a whole, it is evident that the entrance point is on the AB side in by far the great majority of cases, and tends to be opposite the D cell. This may mean either that there is a preferred side for sperm entrance, or that the point of entrance determines as a rule the AB side.
In the previous paper referred to a similar result was observed in the eggs of Cumingia. The I) cell consistently lies opposite the side on which the sperm enters. This occurs where the first cleavage plane coincides with the entrance point, and it follows from the relation between the type of second cleavage and the position of the AB blasto- mere with respect to the entrance point at the 2-cell stage. In Xcrcis the same roult was obtained when the cases of coincidence of the first cleavage and the entrance point are considered. In 28 cases the D formed on the side opposite the entrance point and in five on the same side.
The method followed in these as well as in the former experiments is not well suited to determine whether or not there is in reality a pre- ferred side of the egg at which a spermatozoon is more likely to enter. The eggs, to which a dilute sperm suspension is added, lie between the cover slip and the glass slide (without compression), and the sperma- tozoa, that come into contact with the egg almost immediately, are watched. It is by no means the case that the first spermatozoon, that makes contact with the egg, is the one that enters, but whether this is because the spermatozoon fails to make the proper head-on contact, or whether it has made contact with a side of the egg that is less easily entered cannot be stated. It is true that, with some experience, one can generally tell at once whether a given sperm is likelv to enter or not. If it is about to enter its tail ceases to vibrate and stands out stiffly at right angles to the surface, but if it fails its tail continues to vibrate for some time. ( >nlv those spermatozoa that lie at the edge of the egg can be clearlv observed. Those that enter above or below the rim of the egg cannot be seen. Hence, there is necessarily a certain amount of selection, but this in itself does not seriously affect the situa- tion since the selection of the eggs observed is a random one. It might be possible, by pushing each egg halfway into a small glass tube closed at one end, that has the same diameter as the egg, to determine whether the spermato/oi'in enters only some of the eggs, namely, those with the | >re f erred sj.lc exposed, or whether if they do enter all the eggs tin- entrance is delayed in approximately half of the cases. In this way it mi^ht be possible to find out if there is a preferred side, in other words, whether there is a predetermined dorso-ventral axis.
SOME EXPERIMENTS UPON TEMPERATURE ACCLIMA- TIZATION AND RESPIRATORY METABO- LISM IN FISHES1
F. B. SUMNER AND P. DOUDOROFF (I'roin the Scripts Institution of Oceanography, La Jolla, California)
Introductory, Procedure, etc.
In a preliminary communication from this laboratory (Sumner and Wells, 1935), experiments were described in which fishes of varying temperature history were subjected to a number of lethal agents and anesthetics. The most important conclusions drawn from these ex- periments were stated as follows :
' Fishes which had been kept at widely different temperatures, when tested at these temperatures, were found to differ markedly in the rate of their visible respiratory movements. . . . Other things equal, these are higher at higher temperatures than at lower ones.2 . . . When fishes which had become acclimatized to high and low temperatures, respec- tively, were transferred to a common temperature (usually an inter- mediate one), those from the warmer water displayed a lower respira- tory rate and a lower susceptibility to the anesthetics and lethal agents than those from the colder water. Thus the differences in physio- logical activity which had been originally induced by acclimatization to these differing temperatures were completely reversed when the fishes were brought to a common temperature. After this transfer, it was the former " warm " fishes which had the lower respiratory rate etc. and the former " cold " fishes which had the higher. Fishes kept at intermediate temperatures remained intermediate.
' Acclimatization to a high temperature appears to consist, in part at least, of a change whereby an initial great increase in metabolic rate is followed by a regulative process, with continued sojourn in warm water, and, conversely, acclimatization to a low temperature involves a similar process working in the opposite direction."
Somewhat earlier N. A. Wells (193Sa), from actual determinations of oxygen consumption in Fundulits parvipinnis, had concluded that
1 Contributions from the Scripps Institution of Oceanography, New Series, No. 9.
2 The further statement in the original text relative to higher susceptibility to urethane at higher temperatures has not been substantiated by our later work.
403
404 F. B. SUMNER AND P. DOUDOROFF
" the rate of metabolism of fishes at any given temperature is dependent upon the temperature of the water to which they have been [previ- ously] acclimatized."
That the resistance of a living organism to cyanides and some other poisons varies inversely with the temperature was established by Child for planarians and other invertebrates (Child, 1913). Gcppert (1889) is said to have proposed the view, now generally accepted, that the cyanides act by preventing the tissues from utilizing oxygen.
The occurrence of a certain amount of physiological adaptation in the effects of temperature upon metabolic rate was pointed out by Miss Behre (1'MSi. working upon Planaria dorotocephala. In a discussion of experiments in which the periods of acclimatization ranged from three days to three months, she states: ''Worms tested immediately after they have been put into a higher temperature than that at which they have been living for a shorter or longer time show greater suscepti- bility to cyanide than those which have been living indefinitely at a higher temperature; those tested immediately after they have been put into a lower temperature than that at which they have been living for a shorter or longer time show a lower susceptibility to cyanide than those which have been living indefinitely at the lower temperature."
Our discussion has thus far related to the effect of temperature upon metabolic rate, and the only result of acclimatization thus far considered has consisted in a fall in this rate, following an initial rise, or a rise following an initial fall. These changes in metabolic rate have been detected by observing the relative susceptibility to KCN or certain other poisons. Another type of acclimatization has to do with the effects of subjection to moderately high or low temperatures upon subsequent resistance to heat and cold, when these are used as lethal agents. That this latter class of effects is quite distinct from the former one will appear probable- from facts to be reported below.
Among tin' earlier experiments in acclimatizing vertebrate animals to relatively high temperatures, those of Davenport and C'astle (1895) an- so familiar that a brief rcferemv to them is sufficient. By sub- jecting toad tadpoles for four weeks to 24°-25°, instead of 15°, at which the controls were living, the temperature necessary to produce heat -rigor was raised by 3.2° C.
l.oeb and \Ya.steiieys (1912) conducted some highly interesting ex- periments upon !:itii(lnlns (presumably F. heteroclitus) . These experi- ments were largely concerned with the effects of various salts upon heat resistance. But the authors also sought to determine the maximum temperature which the fishes could tolerate, if transferred suddenly from their natural medium (10° to 14° at the time); "how long it
TEMPERATURE ACCLIMATIZATION IN FISHES 405
takes to immunize the fish against the harmful effects of a sudden trans- fer to a temperature of 35° C"; how long this immunity would last after return to cooler water, and some other matters. Certain of the results of Loeh and Wasteneys will he referred to later. For the most part, their paper does not report the numbers of fishes employed (where stated, they are very small), nor the variability of the individual records, so that the reader is left in doubt as to the statistical certainty of some of the results.
A rather extensive series of experiments on temperature acclima- tization was conducted by Hathaway (1928) upon several species of fresh-water fishes as well as upon toad tadpoles. Hathaway's proce- dure consisted chiefly in keeping the animals for various periods at 10° and 30°, and then determining their mortality rate throughout various periods (1 minute to 24 hours) at temperatures ranging from 30° to 40°. " Continued exposure to high or low temperatures progressively raised or lowered the limit of tolerance of each species," a statement which is abundantly supported by his tables and histograms. No at- tempt was made, however, to relate the period of acclimatization in any quantitative way to the resulting degree of resistance.
The results herein reported are the outcome of experiments con- ducted by the authors throughout about a year past, in the course of which nearly three thousand fishes were employed. Except in one minor experiment, these fishes were gobies of the species Gillichthys uiirabilis Cooper.
Our experimental material was kept in tanks of running sea-water at constant temperatures.3 Tanks were maintained throughout at 10°, 20°, and 30° C., and for part of the time others were maintained at 15° and 25°. Prior to experimental treatment at high or low tempera- tures, all fishes were kept for some weeks or months at the intermediate temperature (20°). In some of our tables and charts, fishes from the 10°, 20°, and 30° tanks are designated as " C," "I," and " W," respectively.
The problems which have concerned us fall into two main groups as indicated in the preliminary discussion: (1) the effects of the sub- jection of fishes to higher or lower temperatures upon their respiratory metabolism; and (2) the effects of this sort of conditioning upon their
3 Except on infrequent occasions, the temperatures were held to within 0.5° of the figures stated, by thermostatic control or otherwise. Owing to failure in the control mechanisms, water in the 10° tank occasionally exceeded or fell below this figure by several degrees. The temperature of the 30° tank was less subject to such perturbations, though it fell some degrees below the required temperature on several occasions. It never rose above this, however, except very early in the course of these experiments.
406 F. B. SUMNER AND P. DOUDOROFF
subsequent resistance to heat or cold, when these were employed as lethal agents.
(1) Some of the relations between this sort of temperature condi- tioning and respiratory metabolism were investigated by N. A. Wells (19350, \935b, 1935c) in this laboratory some years ago, through the direct measurement of oxygen consumption in Fnndulns parvipinnis. In the present studies, no such direct measurements have been made. We have, however, tested some of these relations abundantly in other ways.
For this purpose, fishes of varying temperature history have been subjected to two principal lethal agents: KCN (0.001 M in sea water) and boiled sea water, containing about 5 per cent of the oxygen nor- mally present. Since death from cyanide, as already stated, is due, in part at least, to asphyxiation, the rate of death from either of these lethal agents may be taken as a measure of respiratory metabolism.
In using each of these agents, our tests were of two sorts. The fishes were either (a) tested at the temperature, high or low, to which they had been acclimatized; or (b) they were tested after they had been returned for varying periods to the water of intermediate temperature (20°) from which they had been taken at the commencement of the experiment. In the latter case, the relations were also investigated be- tween the duration of conditioning and the duration of the persistence of the induced effect.
In a limited number of experiments, the fishes of various temperature histories were subjected to the anesthetic urethane. The results from these experiments were less intelligible than those from other lethal agents, and the relation, earlier suggested (Sunnier and Wells, 1935), between metabolic rate and susceptibility to this drug, now seems rather doubtful.
(2) Under the second head, we have chiefly sought to determine (a) the relation between the duration of previous conditioning and the extent of the subsequent resistance to heat (i.e. how high temperatures were lethal to the fishes), and (b) the relation between the duration of conditioning and the persistence of the effect, when the fishes were returned for varying periods to their original temperature.
In tlu>e experiments with lethal heat, our procedure has consisted in planng -mall lots of the fishes under comparison simultaneously in separate screen ra^es, immersed in a common tank. In any given set of experiment-, the temperature at the outset was brought to the same level (most frequently 37.6°), before introduction of the fishes. It was thereafter raised at an average rate of about 0.1° in two minutes. Record was kept of the moment when each fish succumbed, and of the
TEMPERATURE ACCLIMATIZATION IN FISHES 407
temperature of the water (to 0.1° or less) at that moment. Thus, it is evident that the temperatures recorded as lethal and the times recorded as the duration of resistance have no absolute significance. Their sig- nificance is entirely relative to the conditions of these experiments. In most of our experiments, there was little or no overlapping between lots of fishes having different temperature histories. This, although the mean lethal temperatures for such lots might differ by less than one degree.
Some tests were also made with a view to discovering the causes of certain unaccountable differences in our experimental results from lots of fishes which might have been expected to agree completely with one another. In general, none of the suspected causes were found to be responsible for these incongruities. The possible influence of dif- ferences in size, for example, was tested, despite our customary en- deavor to select fishes of the same mean size for two lots under com- parison. As will be pointed out below, size, within the range of dif- ferences which concern us here, could not have been a responsible factor in the matter, although size differences of sufficient magnitude were found to affect the respiratory rate and the rate of death in boiled water.
We likewise tested the possible effects of differences in the recency of feeding, and of minute differences in the concentration of the KCN solution. It was found that recency of feeding — even feeding within 24 hours — did not affect the resistance of the fishes to cyanide, while differences in the concentration of the solution as great as 2 per cent had no recognizable effect.
In all of our experiments, the value to be determined was the rela- tive resistance of a given lot of fishes to the lethal agent or anesthetic which was employed. This resistance was measured by the time which elapsed before each individual " succumbed." It was consequently necessary to adopt a uniform criterion of " succumbing," in other words, a definite end-point which could be regarded as comparable throughout a given series.
The criterion adopted was necessarily different for the different types of experiment. In the case of lethal heat, the matter was simple. The end-point adopted was the time when the fish ceased to make any respiratory movements or other visible movements, spontaneous or induced. This time could ordinarily be recorded with an accuracy of less than half a minute. The figure recorded did not, however, neces- sarily represent the time of actual death, since many of these individuals recovered if transferred promptly to cooler water.
Where extreme cold was the " lethal " agent employed, a satisfactory end-point was more difficult to detect, since ready responsiveness to
408
F. B. SUMNER AND P. DOUDOROFF
sharp mechanical shocks persisted long after all spontaneous movements ceased. Although our criterion was changed somewhat from one ex- periment to another, it was constant within any one experiment, so that some of the results are highly instructive.
In experiments involving asphyxiation (KCX and boiled water) the actual death of the fish was revealed by a more or less prolonged con- vulsive seizure, which was readily recognizable in most cases. More often, this commenced rather dramatically, after some minutes of com- plete quiescence, beginning with a vigorous flexure of the body, and passing into a characteristic series of quivering and twitching movements of the mandible, opercula and fins. The length and intensity of this convulsion were inversely proportional to the temperature, its duration in KCX solution averaging about 6 minutes at 10°, 3 minutes at 20°, and 1 minute at 30°. At the latter temperature indeed, it was fre- quently so brief or so inconspicuous that it could not be observed at all. The time of the commencement of these death throes has been adopted as the end-point in experiments with KCN and boiled water.
TABLE I
Mean times of survival in KCN at various temperatures, after three (lax-- accommodation to these, and relative rates of respiratory metabolism on which KCN curve in Fig. 1 is based. The last figures are the quotients obtained by dividing the reciprocals of the various times by the reciprocal of the time at 10°.
|
Number of Fishes |
Temperature of Acclima- tization (±) |
Actual Tem- perature in Bowls during Experiment |
Times of Survival |
Relative Metabolism |
|
|
Mean |
Range |
||||
|
< |
°C. |
||||
|
20 |
10 |
10.7 |
308.5 |
241-375 |
1.0 |
|
19 |
IS |
15.6 |
184.0 |
153-221 |
1.7 |
|
20 |
20 |
20.4 |
109.2 |
89-139 |
2.8 |
|
17 |
25 |
25.0 |
64.4 |
56-78 |
4.8 |
|
20 |
30 |
29.5 |
42.4 |
34-52 |
7.3 |
In experiments with urethane, choice of an end-point was a more difficult matter, "-specially at low temperatures, a fact which detracts from the value of many of our experiments with this drug.
In addition to records of their relative resistance to lethal agents and to urethane. we have- records of the rate of the respiratory rhythm of many of our lots of fishes. The value of these last, though real, is limited. ( 1 ) owing to the responsiveness of the respiratory rate to various unavoidable sensory stimuli, and (2) owing to the fact that there is another "dimension " to the respiratory movements — depth as
TEMPERATURE ACCLIMATIZATION IN FISHES
409
well as frequency. It is likely that the former may influence the amount of water passing through the gill-chambers as much as the latter, which alone can he recorded.
B.W.
KCN
10" 15 20 25
Temperature
30
FIG. 1. Correlations between temperature and respiratory metabolism. The values for KCN and for boiled water (B. IV.) are obtained in each case by divid- ing the reciprocals of the mean times of death at the various temperatures by tlir reciprocal of the time at 10° (equivalent to dividing mean time of death at 10° by each of the others). The upper curve is based upon measurements of oxygen consumption by N. A. Wells (1935, Table II (means of Experiments 4 and 5). Each of these curves is based upon relative values only. Their separate positions on the vertical scale have no significance.
410
F. B. SUMNER AND P. DOUDOROFF
The more significant results of all these experiments may be sum- marized as follows :
POTASSIUM CYANIDE
(1) Death in KCN was speedy in direct proportion to the tem- perature of the medium, fishes dying far more rapidly at a high tem- perature than at a low one. This fact is altogether in accordance with expectations, since the oxygen requirements of fishes and other poikilo- therms are known to rise with increasing temperature.4 When the reciprocals of the times of death are plotted against temperature, using the value at 10° as unity, a fairly exact logarithmic series is shown. of a form similar to that shown when the known data of oxygen con- sumption arc similarly plotted (Table I. Fig. I).5 It will be seen that the Q10 lies between 2.6 and 2.8.
(2) A low correlation exists between the duration of acclimatiza- tion to high or low temperatures and the degree of resistance to KCN, at those temperatures. When fishes which had been subjected to 30° for a few hours, and for various numbers of days, up to a month or more, were compared in KCX at this temperature, the results were highly variable, though statistical differences of reasonable certainty were evident when all of the results were considered. From Table II
TABLE II
Relation between duration of acclimatization in 30° and resistance to KCN at 1 1n- latter temperature. (Mean times in minutes).
|
1 Jurat ion acclim. |
' hr -5 lirs |
1 da\ |
3 d.i \ - |
5 days |
|
N iimlicr ot fishes |
76 |
52 |
60 |
10 |
|
Length of resistance |
34.67±0.25 |
41.79±0.42 |
44.52±0.53 |
47.90 |
|
1 Mir.ii ion ucclim. |
10-1 1 da: |
22^2 days |
||
|
Number of ti fie |
26 |
42 |
||
|
' li ot resi-i ance . |
50. 85 ±0.67 |
48.36±0.66 |
||
it is evident that a progressive increase of resistance occurred, com- mencing \\itli the briefest periods of acclimatization.
1 It is p<.ssiblr ,,f course, that the results with KCN may be influenced to some extent by tin- effects of temperature on diffusion, or on the "coefficient of dis- tribution," or s, ,nic other physical factor (cf. discussion by Child, 1913), though this could hanll . the phenomena of reversal at a common temperature.
•' Similar cm ° ntin.u the correlation between temperature and me-
tabolism have been published by Krogh (1916), Benedict (1932) and doubtless others.
TEMPERATURE ACCLIMATIZATION IN FISHES
411
This situation may be explained on the supposition that the initial rise in respiratory metabolism, resulting from the abrupt transfer to a higher temperature, is followed by an adjustment, in which the metabolism gradually falls again (Fig. 2). That it never even approxi- mately reaches its former level, at least within the time limits of our experiment, is obvious from an inspection of the table and graphs.
Mm. 300
250
200
150
100
50 30
o
o
eu
3 Q_
OJ O
Mean Endurance at 20
i
30°
.0135
10
34 J 3 5
10
Days in I0°and 30°
o
Days in 20
FIG. 2. Relations between duration of acclimatization in 30° and 10° re- spectively, and resistance to KCN at these temperatures. The 30° curve is based upon data in Table II, the 10° curve upon data in Table III. The line representing "Mean Endurance at 20° ' (102 minutes) is based on all of the fishes (100) for which reliable figures at this temperature are available. Abscissas represent (left) days of acclimatization in 10° and 30° water, and (right) days following return to original temperature (20°). Ordinates represent survival time in minutes.
A similar, though converse, situation was to be expected when fishes were transferred to water of lower temperature instead of higher. An initial fall in respiratory metabolism (increase of resistance to KCN) would occur, followed by a compensatory rise, manifested in diminish- ing resistance. This, indeed, is what we have observed, though the variability of the individual figures is even greater than in the case of the warm water series, and the statistical probability of the successive steps is correspondingly less (Table III). A further difference be-
412
F. B. SUMNER AND P. DOUDOROFF
tween the "warm" ami "cold" scries is the fact that the maximum resistance of the latter (peak of the curve) is not attained as promptly as is the minimum resistance of the former, so that a further rise occurs within the period of our test, hefore the compensatory drop commences (Fig. 2).
(3) If, instead of subjecting the fishes to KCN, while at these temperature extremes (10° and 30°), they were removed from these last and returned to an intermediate temperature before the test with cyanide, it was found that a complete reversal took place, the fishes of 'warm" history now displaying a higher resistance than those of "cold" history, while the 20° fishes remained intermediate as pre- viously. These reversed effects persisted clearly for 3 days after trans- fer to tlu- intermediate temperature, less certainly after 5 and 10 days. (Table IV. Figs. 2, 3.) The re-reversal of the relations which was
TABLE III
Relation between duration of acclimat i/ation in 10°, and resistance to KCN at the latter temperature. (Mean times in minutes).
|
Duration acclim.. |
\ hr-5h hrs |
1 dav |
3 days |
|
Number of fishes |
74 |
53 |
50 |
|
Length ot resistance |
300 8±2 92 |
317 8±? 86 |
302 8 ±4 32 |
|
I Duration acclim. |
5 da vs |
10-11 days |
24-44 days |
|
Number of fishes |
10 |
26 |
33 |
|
Length of resistance |
2963 |
289 1±3 81 |
285 5 ±5 58 |
found at the end of 20 days (Fig. 2) may well be accidental, as are perhaps the exact relations of both of these curves to the " intermediate" (20 ' ) line, which is based upon the mean figure for 100 fishes.
( Vrtain relations of more or less certain significance are to be noted from inspection of Fig. 3. Thus, the curves for the " \V ' (30°) series, without exception, lie above those for the "C (10°) series; the curves oi each comparable pair (equal times in " I ") diverge from one another, with increasing times of acclimatization; the terminals of the curves (with the exception of " I, 2 hrs." in the " \V " series) are arranged in sequence, according to the length of subjection to " 1 ' (20°). and an- arranged in reverse order in the two series. The two curves, for example, are farthest apart, the "1. 10 davs " ones closest together, etc. The- various incongruities in the figure, as a whole, are douhtles, due to the high individual variability shown In- each set of fishes.
TEMPERATURE ACCLIMATIZATION IN FISHES
413
(4) It is probable that vvitbin broad limits no appreciable relation exists between size and susceptibility to KCN poisoning. In two ex- periments, 21 "large" and 17 "small' fisbes were compared, tbe former averaging from 4 to 5 times as beavy as tbe latter. Tbe mean times of succumbing were 101.0 and 103.4 minutes, respectively, tbere
TABLE IV
Persistence of effects of acclimatization to 10° ("C") and 30° ("W"), when followed by sojourns of varying length at intermediate temperature (20°) upon resistance to KCN at the latter temperature. Survival times in minutes. Values having high probability denoted by asterisks. For method of computing probabili- ties see Tippett, 1931, pp. 80-82.
|
Numbers in the two lots |
Duration of acclima- tization |
Time in 20° |
C |
w |
Difference (W— C) |
Probability that difference is significant |
|
29, 24 |
1 day |
2 hrs. |
89.7 |
104.4 |
14.7 |
*0.9999 + |
|
7, 10 |
1 " |
1 day |
90.7 |
107.2 |
16.5 |
*0.9959 |
|
10, 10 |
1 " |
3 days |
109.8 |
109.7 |
-0.1 |
0.5000 |
|
10, 10 |
1 " |
10 " |
100.5 |
101.2 |
0.7 |
0.5000 |
|
5, 5 |
3 days |
2 hrs. |
86.0 |
87.0 |
1.0 |
0.6500 |
|
5, 6 |
3 " |
1 day |
93.0 |
124.8 |
31.8 |
*0.9999 + |
|
15, 15 |
5 days |
2 hrs. |
90.5 |
115.5 |
25.0 |
*0.9999 + |
|
9, 10 |
5 " |
1 day |
90.8 |
121.3 |
30.5 |
*0.9999 + |
|
8, 10 |
5 " |
3 days |
101.4 |
116.0 |
14.6 |
*0.9985 |
|
10, 10 |
5 " |
10 " |
107.2 |
108.4 |
1.2 |
0.6300 |
|
11, 11 |
10-11 days |
2 hrs. |
83.8 |
107.0 |
23.2 |
*0.9999 + |
|
23, 18 |
33-40 d. |
2 hrs. |
84.2 |
123.2 |
39.0 |
*0.9999 + |
|
18, 20 |
36-51" |
1 day |
92.0 |
129.0 |
37.0 |
*0.9999 + |
|
19, 20 |
36-56 " |
3 days |
98.5 |
127.0 |
28.5 |
*0.9999 + |
|
10, 10 |
57 davs |
5 " |
101.7 |
110.9 |
9.2 |
*0.9969 |
|
18, 30 |
31-46 d. |
10 " |
111.9 |
116.7 |
4.8 |
0.8200 |
|
10, 10 |
46 days |
20 " |
110.7 |
103.6 |
-7.1 |
0.9245 |
being, in eacb of tbe experiments, an almost complete overlapping of tbe values.
BOILED SEA WATER
( 1 ) When kept throughout the tests at the various experimental temperatures, the time relations of the deaths in boiled water were very similar to those in KCN though forming a less perfect logarithmic series (Table V, Fig. 1). The values for Q10 are here much less con-
414 p. B. SUMNER AND P. DOUDOROFF
stant, ranging from 2.3 to 3.0. These incongruities may be clue to the considerably smaller number of individuals used.
(2) Death in deoxygenatecl sea water was more than 60 per cent slower, at the same temperatures, than in KCN solution. That this slower rate was not due to traces of oxygen left in our boiled water is known from the fact that considerable increases ()(2 or 3) in the small amount of oxyygen present were not found to have any appreci-
M in 130
W 120
110 100
80
\ a-- .....
L 3 days
C
day 90
o
5 10 30 35 40 45
Days Acclim.
IK.. 3. Acclimatixation for varying periods in "cold" (10°) and "warm" (30°), follo\\rd liy sojourn fur varying periods in "intermediate" temperature 1 i, lYi'iii which the fislies \\i-re originally taken. Abscissas = times of acclima- tixation in ( and //' (days). Ordinates = survival times in KCN solution (minutes). Separate curves are plotted for the various sets differing in their of si pjoiirn in /.
abK- ctTcct. It may be due to the- presence of a considerable reserve of o.xvgen in the tissues, which is rendered unavailable by the- KCN. Or it may be that the latter has some toxic effect, independently of its effect upon oxygen utili/.ation.
(3) ( hving, in part at least, to the much smaller number of fishes used in tin present series, the relations between duration of temperature conditioning and resistance ti> asphyxiation were not so well shown here as in the case of K< X. l''<>r those which had been acclimatized
TEMPERATURE ACCLIMATIZATION IN FISHES
415
for varying periods to 30°, a trend was nevertheless indicated, which agreed with that for KCN. This was not, however, true of the experi- ments at 10°.
(4) In two experiments in which "cold " lots were compared with "warm" or "control" (20°) lots, after transfer to the latter tem- perature, we had the same marked reversal as in the case of cyanide. Here, however, we had only 2 to 4 individuals of each sort.
(5) The size of the fishes, in experiments with boiled water, as contrasted with those in KCN or lethal heat, was found to have a marked influence upon the death-rate. Smaller (i.e. younger) fishes, as might have been expected (for both KCN and boiled water), were the first to succumb. Four experiments were performed, five fishes of
TABLE V
Mean times of survival in boiled sea water, at various temperatures, after 3 days acclimatization to these, and relative rates of metabolism on which "B.W." curve in Fig. 1 is based. For significance of figures for "relative metabolism" see Table I.
|
Number in each lot |
Mean length |
Temperature of acclima- tization |
Actual tem- perature in flasks during |
Times of survival |
Relative Metabolism |
|
|
(cm.) |
(±) |
experiment |
Mean |
Range |
||
|
10 |
10.6 |
10 |
9.8 |
483.7 |
397-619 |
1.0 |
|
10 |
10.2 |
15 |
14.9 |
257.6 |
210-296 |
1.9 |
|
10 |
10.3 |
20 |
19.9 |
179.7 |
140-227 |
2.7 |
|
9 |
10.7 |
25 |
24.8 |
112.0 |
77-148 |
4.3 |
|
10 |
10.3 |
30 |
29.5 |
60.5 |
50- 77 |
8.0 |
each size being used in each of these. In two experiments, in which the mean weight of the large fishes was 2% times that of the small ones, the average duration of life of the former was 15 per cent greater than that of the latter. In two other experiments, in which the large fishes averaged 4 times as heavy as the small ones, the difference was 41 per cent. The means for the 20 large and 20 small, comprised in the entire series were: 297.4 and 238.7 minutes respectively.6 The rapid decline of respiratory metabolism with increasing size in fishes and other cold- blooded animals is now a familiar fact. (Keys, 1931, and Wells, 1935a, have presented data for fishes, based upon experiments con- ducted in this laboratory.)
6 All four experiments yielded large mean differences, while in two of these there was no overlapping of values. But the four sets are not homogeneous enough in respect to temperature, etc. to justify the computation of probabilities.
416
F. B. SUMNER AND P. DOUDOROFF
LETHAL UKAT AND COLD
( 1 ) \\"ith fishes taken originally from a stock tank at 20°. differ- ences in the length of conditioning to 30° led to marked differences in their subsequent resistance to lethal heat. Even a half hour's previous stay at 30° led to a considerable increase in resistance, and this increase continued, though at an ever-slowing rate, up to 10 (perhaps even 30) days, when this particular test was discontinued. After the first day, the rise of the curve was very gradual (Table VI, Figs. 4 and 5).
TAHLK VI
Relation- bet ween length of acclimatizat ion to 30° and resistance to lethal effects of heat, the re-i-t.mce being indicated both by the time and the temperature at which the fishes succumbed. (Including only those experiments in which heat treatment commenced with 37. 6±).
The probabilities relate to the differences between each temperature and the one in the preceding column. High probabilities indicated by asterisks.
|
Time of acclimatization |
0 |
30 inin. |
1 hr. |
i i\ hrs. |
1 5 hrs. |
9 hrs. |
1SJ to 19} hrs. |
|
N timber of fishes. . |
15 |
10 |
5 |
10 |
20 |
10 |
20 |
|
Mean time (min- utes) |
4.9 |
10.6 |
21.5 |
24.5 |
30.1 |
34.4 |
40.9 |
|
Me. m temp |
37.73 |
38.03 |
38.36 |
38.48 |
38.73 |
38.89 |
39.06 |
|
Probabilities |
*0.9999 + |
*0.9890 |
0.8740 |
*0.9999 |
*0.9940 |
*0.9999 + |
|
|
Time of acclimatization |
1 day |
3 days |
10 days |
30 days |
|
Number of fishes |
10 |
10 |
10 |
10 |
|
Mean time (minutes) |
41.5 |
53.6 |
61.0 |
62.2 |
|
Mean temp |
39. 1 1 |
39.32 |
39.62 |
39.68 |
|
Probabilities |
0.9400 |
*0.9975 |
*0.9992 |
0.8300 |
Xot included in the foregoing are three experiments in which the temperature of the "hot " tank at the outset was 38.75° (±), and in which the fi-he-, under comparison (10 of each) had previously been conditioned to 3d for 3 and 39 days respectively. The mean lethal tempcratun > For the two lot.s were 39.50° and 40.11°, respectively, there being no overlapping of the figures for the contrasting lots.
(2) Return to a common temperature of 20" for considerable pe- riods before testing with lethal heat diminished, but did not annul, the
II MI'KRATUKI-: ACCLIMATIZATION IN FIS1IHS
417
effects of previous acclimatization. For example, a sojourn of 23 days at 20° did not suffice to eliminate the difference between lots which had been kept for 46 days at 10° and 30° respectively, while fishes kept for only a single day at these last temperatures retained a considerable part of the resulting difference after 10 days at 20°. On the other
38
0 2
38
0 I 3
9 17
Hours Ace Mm.
10
30
Days Acclim,
FIG. 4. Correlation between previous acclimatization in 30° tank and re- sistance to lethal effects of heat. Data for shorter periods only (up to one day) are plotted in A ; those for both short and long periods are plotted, on different scale, in B. Abscissas = times of acclimatization in hours (or days) ; ordinates = mean temperatures which proved lethal to the various lots.
hand, the effects of 3 days conditioning at these extreme temperatures had nearly or quite disappeared after 30 days at the intermediate one. This last is not in agreement with certain results reported for Funduliis by Loeb and Wasteneys (1912), who state: "The immunity against a temperature of 35° acquired by keeping the fish for two days at 27C is not lost or weakened if the fish are afterwards kept as long as thirty-
418
F. B. SUMNER AND P. DOUDOROFF
three days at a temperature of from 10° to 14°." It may well be that Gillichthys differs from J: undid us in this respect. These authors like- wise state: " The immunity against a temperature of 35° C. is also main-
Mi n. 40
35 30 25
20 15 10
0 2
Hours
17
26
FIG. 5. This is the counterpart of Fig. AA, the present values representing the mean times in which death occurred, instead of the temperatures.
tained if the fish are kept after the two days' exposure to 27° for two weeks at a temperature of 0.4° C." We have not employed such a low temperature in this connection.
TEMPERATURE ACCLIMATIZATION IN FISHES
419
Table VII 7 comprises the chief results of experiments of this class, while Figs. 6 and 7 depict, for one series of tests, the relations between the periods of conditioning in "cold" and 'warm" water (10° and 30°) and the subsequent sojourn at the '''intermediate" temperature (20°). The divergence of the two sets of curves in Fig. 6, and their
TABLE VII
Persistence of effects of acclimatization to 10° ("C") and 30° ("W"), upon heat resistance, when followed by sojourns of varying length at intermediate tem- perature (20°).
The probabilities relate to differences between lethal temperatures for C and W lots. High probabilities indicated by asterisk.
|
Number of |
Period of |
Subse- |
Temp. |
Period of resistance |
Lethal Temp. |
|||
|
fishes in |
acclim. (10° |
quent stay |
Probability |
|||||
|
each lot |
and 30°) |
at 20° |
start |
|||||
|
C |
W |
C |
W |
|||||
|
10 (9 in W) |
6 hrs. |
2( + ) hrs. |
36.0° |
23.0 |
55.8 |
37.38° |
38.56° |
*0.9620 |
|
10 |
1 day |
2\ hrs. |
36.0° |
15.1 |
67.1 |
36.96° |
39.09° |
*0.9999 + |
|
10 |
il J t |
1 day |
36.0° |
28.7 |
61.3 |
37.50° |
39.11° |
*0.9999 + |
|
10 |
1 1 11 |
3 days |
36.0° |
26.7 |
53.3 |
37.58° |
38.68° |
*0.9999 + |
|
10 |
ti a |
10 " |
36.0° |
22.2 |
36.3 |
37.31° |
38.02° |
*0.9997 |
|
15 |
3 days |
30 days |
36.0° |
18.8 |
20.4 |
36.99° |
37.07° |
0.6900 |
|
10 |
5 days |
2 hrs. |
37.0° |
5.0 |
58.3 |
37.30° |
39.27° |
*0.9999 + |
|
10 |
* t n |
1 day |
36.0° |
29.0 |
70.1 |
39.64° |
39.53° |
*0.9999 + |
|
10 |
1 1 u |
3 davs |
36.0° |
13.5 |
64.7 |
36.80° |
39.11° |
*0.9999 + |
|
10 (9 in W) |
n a |
10 " |
36.0° |
18.9 |
47.4 |
37.25° |
38.66° |
*0.9999 + |
|
10 |
34 days |
2( + ) hrs. |
37.0° |
2.5 |
70.1 |
37.20° |
39.89° |
*0.9999 + |
|
10 |
37 " |
1 dav |
36.0° |
5.7 |
80.9 |
36.39° |
40.09° |
*0.9999 + |
|
10 |
11 i i |
3 davs |
36.0° |
7.4 |
76.3 |
36.54° |
39.85° |
*0.9999 + |
|
10 |
32 " |
10 " |
36.0° |
12.7 |
53.7 |
36.82° |
38.90° |
*0.9999 + |
|
10 |
46 days |
23 davs |
36.0° |
14.1 |
37.6 |
36.87° |
38.23° |
*0.9999 + |
|
23 1 |
n tt |
47 " |
36.0° |
24.7 |
27.9 |
37.36° |
37.55° |
0.8850 |
| In this experiment, 10 "controls" which had been kept continuously at 20° gave mean values: time, 18.3 minutes; temperature, 37.07°.
convergence in Fig. 7 represent significant facts, as do the relative positions of all of the single curves in Fig. 7 and some of those in Fig. 6.
(3) Long-time (38-day) conditioning of fishes to a low tempera- ture (10°) made them more resistant to extreme cold (1.0° -) than
7 Corroboration of some of these results was obtained from a number of experi- ments which could not conveniently be included in the table.
420
T 40'
39' 38' 37' 36'
F. B. SUMNER AND P. DOUDOROFF
i * ~. -O
. ; — r^^^EES-----
DtfTT ----------- "[.3 da)s
_ljg.day.s_.
Days Ac
37
cum.
FIG. 6. Same legend as for Fig. 3, except that the present figure relates to heat resistance, the ordinates representing temperatures which proved lethal to the various lots of fishes.
T. r
40° 39° 38° 37° 36°
Odays
Fn,. /. Based u|><.n the same experiments as Fig. 6, but with different treat- ment of the data. In the present case, the subsequent sojourn in "intermediate" (20°) is plotted on the abscissas axis, while separate curves are drawn for the various periods of ] conditioning to "cold" and "warm."
TEMPERATURE ACCLIMATIZATION IN FISHES
421
ones which were transferred to 10° only 3 days previously. Our evi- dence here is reasonably conclusive, though not lending itself readily to quantitative treatment (see pp. 407-8).
(4) Fishes kept alternately at 10°, 20° and 30° for 7 or 8 coni-
TABLE VIII
Effects upon resistance to heat of previous subjection to "cold " (10°), "intermediate"
(20°) and "warm" (30°), in alternation.
|
Temp, of hot tank at start |
Rate of rise |
Number of fishes |
Temperature history |
Period of resistance (minutes) |
Lethal temper- ature |
|
37.6°-37.7° |
0.1 in |
8 |
5 cycles: W(24 hrs.), 1(12 hrs.), |
32.7 |
|
|
13 min. |
C(24hrs.), 1(12 hrs.), etc. |
||||
|
8 |
5 cvcles: C(24 hrs.), 1(12 hrs.), |
46.0 |
|||
|
W(24hrs.), 1(12 hrs.), etc. |
|||||
|
8 |
Control: I (continuously) |
4.5 |
|||
|
37.0°-37.1° |
0.1° in |
10 |
8 cycles: W(12 hrs.), I (12hrs. ), |
29.5 |
38.20° |
|
2.8 min. |
C(12 hrs.), 1(12 hrs.), etc.* |
||||
|
10 |
7 cycles: W(12 hrs.), 1(12 hrs.), |
42.5 |
38.47° |
||
|
C(24hrs.), 1(12 hrs.), etc.f |
|||||
|
20 |
Control: I (continuously) |
6.0 |
37.30° |
||
|
Probability of difference between |
|||||
|
first and second temperature |
|||||
|
low (0.8100); between first and |
|||||
|
third very high (0.9999 + ) |
|||||
|
36.0-36.1 |
0.1° in |
10 |
9 cycles: W(12 hrs.), 1(12 hrs.), |
66.7 |
39.25° |
|
2.1 min. |
C(24hrs.), 1(12 hrs.), etc. t |
||||
|
11 |
Control: I (continuously) |
15.1 |
36.91° |
||
|
Probability of difference very |
|||||
|
high (0.9999 + ) |
|||||
|
37.6 |
0.1 in |
15 |
W (6 days, continuously) |
42.8 |
39.80° |
|
2.0 min. |
15 |
W and I (12-hr, periods alternate- |
|||
|
ly for 12 days) |
43.0 |
39.79° |
|||
|
15 |
W and I (2-day periods alter- |
||||
|
nately for 12 days) |
43.0 |
39.79° |
* Except that last C and W were each 24 hours, t Thus C time = 2 X \V time.
plete cycles were more resistant to heat than ones which had been kept continuously at 20°. This was decisively true, even in some experi- ments in which the " cold " periods were twice as long as the " warm " ones (Table VIII).8 On the other hand, fishes which had been sub-
! A somewhat similar experiment was performed by Loeb and Wasteneys (1912), with results similar to ours.
422 F. B. SUMXER AND P. DOUDOROFF
jected to these alternations of temperature were less resistant to extreme cold (0° -1°) than ones kept continuously at 20°. This statement is based upon four experiments, involving 28 "alternation" fishes and 19 controls. \Yhile the results cannot well he treated quantitatively. for reasons already stated, this general statement admits of little clouht.
(5) When the water temperature was raised gradually, throughout a period of several hours, the tolerance was considerably increased, in comparison with experiments in which the fishes were subjected abruptly to these higher temperatures. Thus when 15 fishes were placed in a tank at 32.2° and the water gradually warmed, the first death occurred when the temperature reached 37.6°. This was 51/: hours after the commencement of the experiment, and 1% hours after the temperature had reached 37.2°. On the other hand, when 19 fishes of the same original stock were placed abruptly in water at stationary temperatures of 36.9° to 37.2°, they all died within a maximum of 17 minutes, the mean time being 10 minutes.
(6) In four -experiments, comparisons were made of the resistance to heat of large and small individuals (totalling 21 of each), having identical temperature history. The former averaged about four times as large, by weight, as the latter. The mean temperatures at which they succumbed averaged: large, 37.79°; small. 37.48°. Comparison of individual figures makes it highly improbable that any relation exists between size and resistance to heat, or at least any which could be revealed without the use of much greater numbers of fishes.9 This is in marked contrast to the consistent differences which were generally to be observed in com] taring fishes of different temperature history.
(7) In a single experiment, the question was tested whether re- sistance to heat was affected by the salt concentration of the water. 1'nlike I .neb and Wastencys (1912), we used for this te-t a fresh water species (Gainbusia affinis). This fish can be acclimatized to 75 per rent sea water (probably to full strength) if the change is made in several steps. In our single experiment, two lots, of 17 and 16 fishes respectively, which had been kept in fresh water and 75 per cent sea water (the latter in this medium for 5 days), were subjected to lethal heat. Tlu- mean temperature of succumbing of the fresh water indi- viduals \\as 37.4u. that of the salt water ones 38.53. If one individual wen- removed from each series, there would be no overlapping between the two. Thus heat was more rapidly fatal to this fresh water species
B Belehradek (1(M5) cit<> contradictory figures as regards the relation of age to heat-resistance in (islirs. M. M. Wells (1914), on the other hand, on the basis of experiments uj ial species of fresh-water fishes, concludes: "Large fish
of a given sp< • more resistant to high temperatures than small fish of the
same species." Wells used .Imciitnts inclas and several species of Notropis.
TEMPERATURE ACCLIMATIZATION IN FISHES 423
/;/ fresh water, than in the unfamiliar medium to which it had been rather rapidly acclimatized.
URETHANE
( 1 ) In comparing the times required for anesthetizing at different temperatures, no consistent results were obtained. The earlier infer- ence (Sumner and Wells, 1935) that the effects of urethane were di- rectly related to the rate of respiratory metabolism does not appear to be well founded.
(2) Nevertheless, when acclimatization at 10° and 30° was fol- lowed by sojourn at a common temperature (20°), tests with urethane, made at the latter temperature, gave consistent results similar to those reported by Sumner and Wells, and agreeing with those reported in the present paper for KCN and boiled water. Fishes of previous
' warm " history were much more resistant than ones of previous " cold " history, and these differences tended to be great in proportion to the duration of the previous acclimatization.
(3) When comparison wras made of fishes (10 each) which had been kept at 30° for 5 hrs., 1 day, 3 days and 34—35 days, a consistent increase in resistance was manifested, the last set requiring nearly three times as long for anesthetization as the first.
(4) No consistent results were obtained when similar tests were made at 10°.
This last fact is due, in part, to the difficulty in finding a satisfactory " end-point " for urethane anesthesia at low temperatures. For this and other reasons, we abandoned the plan to make more extensive tests with this drug. While the results stated in paragraphs 2 and 3 of this section appear to represent actual physiological differences, we are not disposed to attempt any interpretation of these at present.
RESPIRATORY MOVEMENTS
( 1 ) The frequency of respiratory movements, based upon 45 " cold ': (10°) fishes, 78 "intermediate"10 (20°) ones, and 44 "warm" ones, was 26.4 ± 0.61, 73.7 ± 1.15, and 74.8 ± 1.95, respectively, per minute. The figures upon which these averages are based are restricted to experi- ments in which fishes at all three temperatures were available for simul- taneous comparison. It is unlikely that the ' warm " and " inter- mediate " fishes differed significantly (Table IX).
10 This number included many of the " warm " and " cold " fishes, which were tested at 20°, before subjection to the other temperatures.
424
F. B. SUMNER AND P. DOUDOROFF
(2) Transfer from 20° to 10° resulted in a mean fall in respiratory rate of 39.7 ± 2.02 per minute (mean of 39 fishes). After return to 20°, this was followed, in the 37 fishes thus tested,11 by a mean rise of 32.5 ± 1.34 per minute.
(3) Transfer from 20° to 30° resulted, quite unexpectedly, in no consistent rise in respiratory rate. In seven experiments (3 fishes each), in which counts were made of the breathing of fishes that had been transferred to the 30° tank from 2 hours to 4 days previously,
TABLE IX Mean rates of respiratory movement at 10° ("C"), 20° ("I") and 30° ("W")
|
Duration acclim. in C and W |
Temp, history |
Number of fishes |
Mean length (cm.) |
Resp. per minute |
|
C |
6 |
12.7 |
36.0 |
|
|
3-4 Mrs. |
I |
18 |
12.7 |
73.5 |
|
W |
6 |
12.8 |
86.7 |
|
|
C |
12 |
13.1 |
27.5 |
|
|
18-24 hrs. |
I |
30 |
13.0 |
73.2 |
|
W |
12 |
13.3 |
70.8 |
|
|
C |
9 |
13.1 |
26.1 |
|
|
2 days |
1 W |
24 8 |
13.0 12.7 |
72.6 74.8 |
|
C |
12 |
13.2 |
22.4 |
|
|
10-11 days |
I W |
12 12 |
13.0 13.2 |
76.6 81.2 |
|
C |
12 |
13.7 |
29.4 |
|
|
33-37 days |
I W |
12 12 |
14.3 13.9 |
74.4 72.5 |
|
Means (omitting 3-4 horn |
C I W |
45 78 44 |
13.29 13.20 13.33 |
26.4±0.61 73. 7 ±1.1 5 74.8±1.9S |
the mean rise in respiratory rate was a nominal one (1 per minute), then' being actually more negative cases than positive ones. On the other hand, there was a large and consistent fall when 30° fishes were returned to 20°, the average difference for 28 fishes being 33.3 ± 2.50 per minute. These two sets of facts are difficult to reconcile with one another.
(4) \Yhen fishes of 'warm" and "cold" history (28 of each)
1 Mostly different individuals from those covered by the statement im- mediately preceding.
TEMPERATURE ACCLIMATIZATION IN FISHES 425
were compared 15 minutes to 2 hours after transfer to 20°, the latter gave a distinctly higher rate, this difference averaging 15.9 ± 2.73 per minute. Here \ve had the same reversal of relation as was found with KCN and boiled water.
(5) The rate of respiratory movements was dependent to some ex- tent on size. In each of 4 experiments, 5 " small " and 5 " large " fishes were used, the temperatures being 15.1° to 15.5°. The 20 small ones, having a mean weight of about 34 grams, gave a mean rate of 49.1 respirations per minute, as compared with 26.9, for the 20 large fishes, which averaged about 73 grams. In these four experiments, only a single case of overlapping occurred between the two contrasted lots.
DISCUSSION
From the various quantitative relations revealed in our tables and curves, it is conceivable that definite conclusions might be drawn re- garding the nature of the various chemical reactions concerned. The present writers do not feel disposed, or indeed qualified, to undertake any such general interpretations of our results. It may, however, be instructive to point out analogies between certain features of these results and some other biological phenomena.
The situation depicted in Fig. 2 seems to find an instructive counter- part in the field of sensory physiology. The intensity of a sensation rises rather abruptly to a maximum, following a stimulus, and then falls again, even while this stimulus is continued. The crossing of the two curves in our figure, following the return of the two sets of fishes to a common, intermediate temperature, recalls an old and familiar experi- ment in sense physiology. We refer to the one in which the subject's two hands are dipped into hot and cold water, respectively, and then, after a few minutes, they are removed and dipped simultaneously into water of medium temperature. One experiences a reversal of his temperature perceptions quite parallel to the reversal in metabolic rate which is manifested by the fishes after a similar transfer. Such an analogy is not, of course, to be offered as a substitute for an explanation in physico-chemical terms. But in the absence of a satisfactory ex- planation of the latter type, it may have suggestive value. Indeed, it seems likely that this analogy is based upon fundamental similarities in the irritability of quite different sorts of living matter.
The relations shown in Fig. 1 are representative of a wide range of phenomena which have to do with the effects of temperature upon the rates of vital processes. (Cf. Krogh, 1916; Belehradek, 1935.) The vast and well-tilled field of " temperature coefficients " (cf. Beleh-
426 F. B. SUMNER AND P. DOUDOROFF
raclek, 1935, Heilbrunn. 1937) is one which \vc have no intention to discuss.
Our curves (Figs. 4 and 5 ), are quite definite in their trend, though we doubt whether they furnish anything very distinctive by which the underlying chemical or physical processes could be identified. Perhaps others might find this possible. It hardly seems likely, however, that such a rapid acclimatization to heat could result from the production of fats having successively higher melting-point (Heilbrunn. 1937; Belehradek, 1935). though we should not be warranted in denying this possibility. Biochemical processes in the field of sense physiology are known which are even more rapid than this, e.g. light- and dark- adaptation in the human eye (Hecht. 1920, Tansley. 1931 ). The strik- ing similarity between our curves (Figs. 4 and 5) and some of those offered by the latter writer perhaps deserves mention. Tansley calls attention to the close parallel between the rate of a sensory change, de- termined subjectively, and a measured biochemical change (restoration of visual purple), which is presumed to be intimately associated with this. Some future physiologist may be able to prove that heat-acclima- tization rests upon an equally simple chemical transformation.
Regarding the much-discussed question of the cause of death at (moderately) high temperatures, we can make no contribution of a positive nature. We can, however, offer reasons for doubting the adequacy of one of the agencies which have been proposed, namely a fatal rise in the rate of respirator}- metabolism. \Ye have already called attention to the difference between the effect of temperature con- ditioning upon resistance to heat and its effect upon resistance to as- phyxiation by boiled water or by cyanide.
The former effect was of far greater magnitude than the latter. Resistance to heat, as measured by survival time, increased eight-fold in the course of one day, and continued to increase quite appreciably for some days further. Resistance to KC'X. measured in the same way, increased about 50 per cent in the first day, and any further increase \vas only detectable statistically. Increase of resistance to asphyxiation in boiled water was even less obvious than in the case of KCX, though this last fact may have been due to the smaller numbers of fishes which we employed.
There were equally striking differences in the persistence of the effects ot" heat < < .nditioning upon resistance to heat and to asphyxiation. The effect upon heat resistance was very much more lasting, after withdrawal of the original temperature conditions, than was the effect upon resistance to I\('X. ( I'oiled water was not tested in this con- nection.)
TEMPERATURE ACCLIMATIZATION IN FISHES 427
Still another difference of interest, already mentioned, concerned the relation of size to susceptibility to these lethal agents. As regards heat resistance, no difference was found, within rather broad limits. In boiled water, on the other hand, the smaller fishes rather uniformly died earlier, as might have been expected. Quite unexpectedly, how- ever, no such difference was found in experiments with KCN. For this we can offer no present explanation.
It would appear, therefore, that there are at least two distinct proc- esses involved in acclimatization to a higher temperature : ( 1 ) a regu- lative decline in the initially much increased respiratory metabolism ; and (2) another change of an unknown nature (either physical or chemical), by which the threshold of susceptibility to destructive tem- perature is raised.
SUMMARY
Fishes (GiUichthys inirabilis) having different temperature histories were subjected to KCN, to boiled sea water, and to temperatures high or low enough to be lethal. Urethane, an anesthetic, was used in a limited number of experiments. The time of death (or of anesthetiza- tion) was recorded for each of the fishes (nearly 3,000). Counts of respiration were made in a considerable number of cases.
Death in both KCN and in boiled water was speedy in proportion to the temperature, the reciprocals of the times of death forming an approximate logarithmic series when plotted against temperature. The temperatures used were 10°, 15°, 20°, 25° and 30°.
In the case of KCN, a low correlation was found to exist between the duration of previous acclimatization to high or low temperatures and the degree of resistance to KCN at those temperatures. The ab- rupt decrease of resistance (increase in metabolic rate), resulting from transfer to a higher temperature, was followed by a slight increase of resistance which continued for several days. On the other hand, the abrupt increase of resistance (decrease in metabolic rate), resulting from transfer to a lowrer temperature, was followed by a slight decrease which continued for several days.
In boiled water, this compensatory trend was only shown after trans- fer to higher temperatures.
When fishes which had been acclimatized to high (30°) and low (10°) temperatures were tested in KCN, boiled water or urethane at 20°, there was a consistent reversal in the resistance of the respective lots, the 30° fishes now being much more resistant than the 10° ones. These relations continued for several days after transfer to the inter- mediate temperature.
428 F. B. SUMNER AND P. DOUDOROFF
Death in de-oxygenated sea water was more than 60 per cent slower, at the same temperatures, than in KCX solution.
Acclimatization at 30°, even for brief periods, resulted in a marked resistance to the lethal effects of heat. This result was already con- spicuous, after a halt' hour at 30°, and increased, at a diminishing rate up to 10 days or more.
Return to a common temperature of fishes having " cold "' and ' warm " histories diminished, hut did not annul, the effects of previous acclimatization upon heat resistance. Pronounced differences were noted in one experiment, after 23 days at the intermediate temperature. Interesting relations were pointed out between the period of conditioning and the persistence of the effects.
Acclimati/ation at 10" increased resistance to lethal cold (1.0°--).
Fishes kept for alternating periods at 10°, 20°, and 30° (even when the 10° periods were twice as long as the 30° ones) were more resistant to lethal heat than fishes kept continuously at 20°. Such fishes were, however, less resistant to lethal cold than the controls.
The rate of the visible respiratory movements was much greater at 20° and 30° than at 10°. There was no certain difference, however, between the rates at the first two temperatures.
Transfer of fishes of " cold " and ' warm " history to an inter- mediate temperature resulted in the same reversal in relative respiratory rates as was found in the case of resistance to KCX, etc.
The size of the fish appears to have little or no influence upon its Tesi-tance to lethal heat or to cvanide poisoning. On the other hand, it was found that smaller fishes had a more rapid respiratory rhythm, and succumbed more rapidlv in boiled water than larger ones.
Acclimatization to higher or lower temperatures brings about two classes of effects which seem to be largely distinct from one another. ( )ne of these concerns increased resistance to the harmful effects of heat and cold, when these are used as lethal agents. The second of these concerns changes in respiratory metabolism, and therefore in re- sistance to oxygen lack, including KCN poisoning. Reasons are given for believing that the former effect is not dependent upon the latter.
LITERATURE CITED
BKIIKI, Kiixnk II., 1<)1S. An experimental study of acclimation t» temperature
in I'lanaiia dorotocephala. H'ml. Hull., 35: 277-317. BKLEHRAUKK. J., 1'^.vx Temperature and living matter. Protoplasma-Mono-
.</;•<;/>///<•». 8: _'77 pp. Berlin. Gebruder Borntracger. BENEDICT, F. (i.. l'M2. The physiology of large reptiles. C'amcyic lust, of U'asli-
inyton, r»h. No. 425; 539 pp. Ciin.ii, C. M., 1(>13. Studies on the dynamics of morphogenesis and inheritance in
experimental reproduction. V '. Jour, ll.rfcr. Zool., 14: 153-206.
TEMPERATURE ACCLIMATIZATION IN FISHES 429
DAVENPORT, C. B., AND W. E. CASTLE, 1895. Studies in morphogenesis, III. On
the acclimatization of organisms to high temperatures. Arch. {. Entvj.,
2: 227-249. GEPPERT, J., 1889. Ueber das Wesen dcr Blausaurevergiftung. Zcitschr. /. klin.
Med., vol. 15.
HATHAWAY, E. S., 1928. Quantitative study of the changes produced by acclima- tization in the tolerance of high temperatures by fishes and amphibians.
Bull. Bureau of Fisheries, 43: 169-192. HECHT, S., 1920. The dark adaptation of the human eye. Jour. Gen. Physiol., 2:
499-517. HEILBRUNN, L. V., 1937. An Outline of General Physiology. Philadelphia.
W. B. Saunders Co., 603 pp. KEYS, A. B., 1931. A study of the selective action of decreased salinity and of
asphyxiation on the Pacific killifish, Fundulus parvipinnis. Bull. Scripps
Inst. (Tech Ser.), 2: 417-490. KROGH, A., 1916. The Respiratory Exchange in Animals and Man. London.
Longmans, Green & Co., 173 pp. LOEB, J. AND H. WASTENEYS, 1912. On the adaptation of fish (Fundulus) to
higher temperatures. Jour. Expcr. ZooL, 12: 543-557. SUMNER, F. B. AND N. A. WELLS, 1935. Some relations between respiratory
metabolism in fishes and susceptibility to certain anesthetics and lethal
agents. Biol Bull, 69: 368-378. TANSLEY, KATHARINE, 1931. The regeneration of visual purple: its relation to
dark adaptation 'and night blindness. Jour. Physiol. (Cambridge), 71:
442-458. TIPPETT, L. H. C., 1931. The Methods of Statistics. London. Williams and
Norgate, 222 pp. WELLS, M. M., 1914. Resistance and reactions of fishes to temperature. Trans.
III. A cad. Set., 7: 48-59.
WELLS, N. A., 1935o. The influence of temperature upon the respiratory meta- bolism of the Pacific killifish, Fundulus parvipinnis. Physiol. ZooL, 8:
196-227. WELLS, N. A., 1935/>. Variations in the respiratory metabolism of the Pacific
killifish, Fundulus parvipinnis due to size, season, and continued constant
temperature. Physiol. ZooL, 8: 318-336. WELLS, N. A., 1935r. Change in rate of respiratory metabolism in a teleost fish
induced by acclimatization to high and low temperature. Biol. Bull., 69:
361-367.
INTERMEDIATE-WINGED . \PHIDS AND THE TIME-OF- DETEK M I X. \TIOX THEORY '
KARL A. STILES COE COLLEGE, CEDAR RAPIDS, IOWA
Transplantation experiments by Spemann (1927), Mangold (1928), and many other- have shown that the destinies of different parts of vertebrate' embryos are fixed at different times in the1 embryogeny. The same technique is not readily open to the investigator of invertebrate development, but it has been assumed that there is a similar succession of embryonic segregations in these groups. One important genetic the- ory, that of Goldschmidt (1931). concerning the nature of intersexes, .-tales that thev begin development in the manner characteristic of one sex, that a physiological change constituting a turning point occurs later, and that subsequent embryogeny is characteristic of the other sex. An organ which is determined before the turning point is reached will be like that of the former sex; one which is determined after the turning point will be like that of the latter sex. An individual produced as a result of the change postulated above is a time-mosaic of male parts and female parts whose- embryonic segregations occur at different times in development. This concept of development has also been applied to intermediates in general and has been termed the time-of-dctermination theory (Shull, 1930). Since intermediates may conceivably be pro- duced without any time intervals between the determination of their different structure.-, it is desirable to discover whether such a succession of events doe- occur.
Intel-mediates of such a nature that they may be used to test the applicability of the time-of -determination theory an- not common. Ihit when in a species of organism there exist two forms of individuals built on the -ame general structural plan, but differing with respect to recog- ni/ably homologous parts, definite information regarding the time of determination of the differential parts can be gained, provided each of tlie^e form- can be converted by stages into the other. Previous to Sbuir.s H''30) work on aphids. efforts to obtain this information have dealt almo-t excln-ivelv with intersexes which are the product of bi- sexual reproduction. The genetic composition of such intersexes is presumably variable. Intermediate aphids parthenogenctically produced have the advantage of allowing only a minimum of opportunity for the
1 Contribution !'r<>m the IV|i;u-tmcnt of Zoology, University of Michigan.
430
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 431
embryonic process to be influenced by genetical variation, and therefore insuring that genetic and developmental factors will not be confused in the results.
The test of serial embryonic segregation (embryonic determination) lies in obtaining intermediates by an approach from opposite extremes. Shull made such a test on aphids which could be changed from partheno- genetic to gamic, and from gamic to parthenogenetic. This change from one form to the other results in the production of many aphids which are structurally intermediate between the gamic and parthenogenetic. If the time-of -determination theory be applicable to the intermediates produced, the order in which the structures change should be the same in both ; but, since in one of the changes a given organ should appear, while in the opposite change that organ should be lost, the combination of characters in the intermediates produced should be quite different in the two cases. In fact, they should be complementary in the complex of organs concerned.
In Shull's early experiments the combination of characters in the aphids obtained by this method were not complementary; the order of the change was apparently reversed, while the development of the inter- mediate characters was identical regardless of the direction of change. To explain these results which appeared to oppose the time-of-determi- nation theory, Shull discusses the possibility of a physiological level hypothesis as an alternative. According to the physiological level theory, the degree of intermediacy might depend upon the state in which some physiological property existed at some critical time. If all the differentiating structures responded equally to changes in the level of this physiological process, the question of the order of determination would not arise. All organs would appear to be segregated at the same time, or during the same part of the embryogeny. Since, however, one structure might respond to a small increase of a hormone, for example, so that it became changed, while another less susceptible does not, there develops a situation in which a difference in the time of determination of these structures may be suspected, although such a difference were non-existent. This explanation is opposed to the theory that the time of embryonic segregation determines the combination of characters in the intermediate individuals. Shull's (1933) final study of a hetero- geneous group of intermediate aphids of the gamic-parthenogenetic types showed that about three-fourths fell into categories to which the time-of-determination theory was applicable, and nearly one-fourth showed irregularities which could not be made to harmonize with this theory, or which required shifts in the order of determination of the several organs studied.
432 KARL A. STILES
Shull (1937) reports a different type of intermediacy. one in which the intermediates are structurally between the winged and wingless parthenogenetic forms. The intermediates reported in the present paper are of the winged-wingless type. In fact, the intermediates discussed here were a part of the (>.15J reported by Shull (1937).
Shull (1937), after discussing many different possible hypotheses which might account for intermediate-winged aphids, states that whether intermediates arc in any way dependent on the direct ion of some physio- logical change could not be judged from the experimental methods of producing them. Then he adds that analysis of the characters of inter- mediates will IK- needed to answer this question. This investigation represents an effort to make a more thorough study than one involving only external characters to which Shull (1937) confined his studies. There are many limitations to this method, for the possible internal development of an organ before there is any external evidence may be of great importance in explaining the developmental basis for inter- mediate-winged aphids.
The structures involved in intermediate-winged aphids are more minute than in the gamic-parthenogenetic intermediates (Shull, 1930) which were examined internally by gross dissection. Accordingly, the winged-wingles> intermediate aphids of this investigation, after their external organs were rated, were prepared for histological and cyto- logical studies of the internal structures, thus making possible a thor- ough analysis of the whole intermediate complex.
The analysis of the intermediates so studied provides the subject matter of this paper, which records attempts to test not only the relative merits of the time-of-determination and physiological level theories as applied to intermediate winged aphids, but to determine the order of embryonic segregation if the time-of-determination theory proved ap- plicable; and to consider other related developmental phenomena. The author wishes to express here his great indebtedness to Processor A. Franklin Shull under whose direction this work was done.
TIIF. STOCKS OF INTERMEDIATES
All aphids used in this study were from stocks which Professor A. Franklin Shull had been keeping for several years under laboratory conditions. They were from two different strains of the species Macrosiphum solanifolii, which have been designated the 1()23 and W31 (Shull. 1('37) stocks. Shnll ( I'M/) describes in detail the experimental methods used in the production of winged-wingless parthenogenetic male a
THEORY OF EMBRYONIC DETERMINATION IN APHTDS 433
EXTERNAL AND INTERNAL DIFFERENTIATING CHARACTERS
A complete description of the external characters of typical winded and wingless forms of Macrosiphum solanifolii has been given by Patch (1915). The winged female differs from the wingless in possessing wings, three ocelli on the head, wing muscles in the thorax, extra sen- soria on the third segment of the antenme (4—6 in the wingless, 15-18 in the winged), and dark color in the third segment of the antennae. The first three organs named proved of the most critical value for this study.
Internal differences between the winged and wingless are, for the most part, correlated in some way with the external distinguishing char- acters. The wing muscles of the winged individuals are striated mus- cles, both vertical and horizontal in position. The protocerebral lobes of the brain to which the ocellar nerves are connected in the winged type show greater development than in the wingless forms. No qualita- tive or quantitative differences in the " symbiotic organ " were observed. There is a practical difficulty involved in making quantitative compari- sons as the number of " symbionts " gradually decreases with the age of an aphid (Uichanco, 1924).
TECHNIQUE
Toto mounts of intermediate-winged aphids were of some value in comparing the relative development of the median ocellus to the lateral ocelli. Of a variety of methods used for this toto mount work, the best proved to be fixation in Bourn's fluid, dehydration with ethyl alco- hol, clearing with xylol and mounting in dammar.
Histological and cytological studies which involve quantitative con- siderations necessitate a technique which results in almost faultless preparations. If the tissue is to be satisfactory there must be no tearing, shattering or distortion. A great many techniques with many variations were tried in an effort to obtain the most satisfactory method. None of the techniques proved satisfactory until the n-butyl method for animal tissues described by Stiles (1934) was used for dehydration and in- filtration. Legs, wings, and antennae were snipped off with DeWecker eye scissors to facilitate handling and to improve sectioning of the aphids. The sections were cut 7 micra and stained with Ehrlich's acid hematoxylin as a nuclear stain and erythrosin for the cytoplasm.
DESCRIPTION OF INTERMEDIATES General Anatomy
A diagnosis of intermediacy of the winged-wingless aphids has gen- erally been based on external features and us a result intermediates of
434 KARL A. STILES
this type have been defined as aphids that have started to develop wing- but have failed to complete such development. That this is not by itself a diagnostic feature of intermediacy is shown by the fact that there is histological evidence of internal intermediacy without external wings. Nevertheless, the investigation described in this paper is limited almost exclusively to intermediates which could be recognized by their wing characters, because of the enormous amount of labor involved in making histological preparations to identify intermediacy from the internal organs.
The wing character in intermediates is very variable. The partially developed wings in an intermediate may vary in development from a barely perceptible roughness in the thoracic region to .95 of normal development, the latter type being very unusual. Drooping wings was a character common to all the seven species of winged intermediates described by Baker and Turner (1916). and they considered this due to a lack of supporting muscles. While many wings examined by the writer were badly crumpled and misshapen, there was no decided tendency to droop; and even if there had been, it is doubtful if it could have been explained on the basis of a lack of muscle development as many intermediates have muscles well developed in proportion to the amount of wing. Unusual cases were those in which the wings were about one-half developed and would have shown muscle through the cuticula of the dorsal thorax but did not because of a degeneration (to be discussed later) which destroyed the muscle. Fat globules could be seen through the cuticula where muscles would normally have been seen.
Antennal color may be like that of the winged or wingless, or it may be some grade between these extremes (Tables 1 to IX). The sensoria. although variable, are often no greater in number than in the wingless,
EXPLANATION OF PLATE I
FIG. 1. Photomicrograph of a transverse section of the head of aphid 1103 d (4), showing on the left side considerable development internally of the ocellus before there is any thickening of the cuticula to form a lens (X 134).
Fir,. 2. Photomicrograph of a transverse section of the mesothorax of P W 48 (1), showing the wing muscle of a fully-winged aphid (X 105).
FIG. 3. Photomicrograph of a transverse section of the mesothorax of aphid 1087 g (2), showing the reduced wing muscle of an intermediate (X 117).
FIG. 4. Photomicrograph of a transverse section of the mesothorax of P 41 (1), a fourth instar winged aphid (X 88).
FIG. 5. Photomicrograph of a transverse section of the mesothorax of P 25 (2), a fourth instar intermediate aphid showing degenerating wing muscle (X95).
A photomicrograph takc'ii with oil immersion lenses to show the de- tail of degenerating wing muscle of a fourth instar aphid, P 25 (3). The light- colored tissue in the upper part of the picture is normal muscle. Note especially the dark spherical degenerating cells between the muscle masses (X 917).
PLATE I
436 KARL V STILES
and uncommonly approach the- high degre'- of development in the winged (Tables I to IX).
Rudimentary ocelli are frequently jiresent in various degrees of formation. Then- are mum cases in which no ocelli can he seen ex- ternally with either a dis-ection or compound microscope, hut histo- logical preparations gave positive evidence of a partly formed structure internally ( Fig. 1 t.
The histological structure of the wing muscle in imerniediate aphids may he characteristic of that in the typical winged or it may he- de- generate in varying degrees, \\ing muscles ( Fig. 2) which in the typical a]>hid occupy so much of the space in tin- thorax, in the inter- mediate ("Fig. 3) are usually less extensive. Degeneration of larval muscle^ resulting from histolysis in insects with a complete metamor- phosjs is well known. Hut in grou|)s with a paurometabolous i incom- ]ilete) metamorphosis there is usually a continuous transformation of tlu- internal organs from the immalup' stages to the1 adult without his- tolvsis of muscle or other internal structures. \Yith these well-estab- lished facts in mind, one would not expect to find degenerating wing muscles in intermediate aphids unless it developed that they possessed marked physiological differences trom normal tvpes. However, nearly one-half of the intermediates studied showed wing muscle with some degree of degcnerat ion. This degeneration e.\]iresses itself in conditions which range from almost complete dissolution of the wing muscle to muscle that can he told from normal only with difficult v. The histology of this degenerating wing muscle is in some respects unlike that which has heen described for the histolysis of the wing muscle of the queen ant. I.ttsiits nii/cr. by Janet ( 1907). The degeneration in the ant wing mu-cle takes place soon after the nuptial flight. At the beginning of histolysis, striations may be seen in the muscle but they become faint and finally disappear. Then the sarcostyles undergo disintegration but the nuclei persist for some time. The story of histolysis in wing mus- cles of the intermediate-winged aphid differs most from that of the ant in that embryonic muscles art' involved.
The earliest stages in which the liistoly^js has been observed in the thoracic muscle of intermediate-winged aphids is the fourth instar. At tin in the normal metamorphosis the wing muscle has about com-
pleie.1 its development and the muscle generally exhibits the character of that of the ima:
1 >mparisoii of a fourth instar with normal muscle (Fig. 4) and a fourth instar which shows degenerating muscle (Fig. 5) will indicate some of the general differences to lie found. The intermediate nature of the muscle in Fig. 5 is indicaie.l by the small amount. Pxith aphids
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 437
were prepared l>y a similar technic and stained \vitli Fhrlich's hema- toxylin. The degenerate muscle in Fig. 5 is distinctly basophilic whereas the normal muscle in Fig. 4 is decidedly eosinophilic. This basophilic nature of muscle is characteristic of the periphery of the developing muscle mass of the1 earlier instars. This might suggest embryonic rather than degenerating muscle, but in normal developing muscle, while the periphery of a muscle where mitoses are taking place most rapidly may he basophilic. the median portion which has earlier become differ- entiated is eosinophilic. So in embryonic muscle the differentiation is from the center of the muscle mass to the periphery. There are, how- ever, instances in which the degeneration of muscle is slight and these simulate somewhat the developing wing muscle of early instars.
Peripheral degeneration is interpreted as a result of the greater sus- ceptibility to environmental influences which embryonic cells may show. Mottram (1913j with radiation has shown that Ascoris eggs are eight times as susceptible during division as during the resting stage. That is. there are eight times as many deaths following an exposure made during mitosis as during the later period. It has also been observed that cancer tissue is much more susceptible to radiation when it is grow- ing rapidly than when it is nearly stationary. The last part of the aphid wing muscle to remain embryonic is the periphery and if degen- eration of wing muscles, as the author believes, is due to an environ- mental influence, the peripheral tissue would be most affected. Figure 6 is a high power photomicrograph of the wing muscle in a fourth in- star intermediate-winged aphid showing something of the more detailed structure of unusually degenerate muscles. The very dark spherical cells, lacking uniformity in size, which are to be seen in the spaces be- tween the masses of muscles, are in advanced stages of cell degeneration. These cells in the advanced stages of degeneration seen in intramuscular spaces were doubtless at the periphery of the muscles shown and were least developed when the degenerating influence became effective. It may well be that they were more susceptible because of their extremely embryonic character. As is seen, these cells stain intensely with hema- toxylin. The small amount of light colored muscle in the upper part of the picture is normal tissue which had reached the definitive stage and was unaffected by whatever influence caused the embryonic cells to degenerate. It was noted that not only mature wing muscles, but other definitive muscles of the body do not seem to suffer degeneration even though there may be marked histolysis of the embryonic muscles present.
Figure 6, in which histolysis is most active, is favorable material for studying some of the details of the process of degeneration. A loose-
438
KARL A. ST1I.KS
ness of the muscle tissue is apparent, distinct outlines of muscle cells may be seen which normally would be so closely crowded together that they could not be delimited fn>m one another. In a given mass of degenerating muscle the tissue least affected is that farthest from the periphery.
Adult intermediate-winded aphids do not otten show as much de- generation of wing muscles as is illustrated by Fig. 7. The wing mus- cles in Fig. 7 have suffered degeneration to the point where only un- striated loosely fibrous >tr:mds of tissue crowded with irregular, densely staining pyknotic nuclei remain. Kardccu ( 1900) points out that nuclei with a .strong aftinitv tor stain are characteristic of degenerating muscle
*
8
FIGS. 7-8
]•']>,. 7. Photomicrograph <>f a transverse section »\ the nie^>thorax »i aphid JJn e, an intermediate with .(> \vin.u~ <K \ rlnpnu-nt. Tlic vertical \vin»- iiui>ck> arr loose, til>r'iiis. ini^triaifd strands oi ti--iK', an example- «\ extrrnic mu^ck- drt;eii-
in the adult { / 115).
I-'IL. S. Photomicrograph «i a trau>vi-r>r section of the nieMithorax of ajihid u (2), with .22 wint; develMpment and h;M"l' inieall v niTinal \vinu muscle
(X 128).
in tin- pig embryo. Associated with the degenerate vertical muscle on the right side may be seen a group of either degenerate muscle cells or phagocvtr-> distended with muscle di-bris. Some of the larger of these spherical bodio are muscle cells undergoing dissolution, but sonn- of the smaller have the characteristics of leucocytes which are idled with di- gesting mus< le fragments.
The longitudinal as well as the vertical muscles are involved in the degeneration process. The longitudinal muscles have so completely degenerated in big. 7 that then- are only a few degenerate fragments left. The question whether a wing muscle degenerates or not depends,
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 439
as has been pointed out, chic-fly on its stage of development when condi- tions for degeneration obtain. The importance of wing muscle degen- eration in intermediate-winged aphids in its relation to the time of em- bryonic determination of organs will be considered in another part of this paper.
It would be logical to examine the developing wings and ocelli in nymphal winged intermediate aphids with degenerating muscles to see if they experience any histolysis. If reference is again made to Fig. 5, the wing will be found to be reduced in size for a fourth instar and there are indications of degeneration. Cells of hypodermis of the lower and upper side of the wing, instead of being elongated as in the normal developing wing, tend to become spherical in some instances, and show a marked affinity for hematoxylin as do the degenerating muscle cells. Other disintegrating cells do not become spherical, but the cytoplasmic part of the cell can plainly be seen to be breaking down. The wing tissue in process of degeneration exhibits a marked affinity for hema- toxylin. Degenerate cells are scattered in the space between the upper and lower sides of the wing bud.
From this study of the internal microscopic anatomy of intermediate aphids it is seen that there are two distinct anatomical situations, one in which the organs concerned with intermediacy are normal in his- tology, the other in which at least the wing muscle and developing wing bud show more or less degeneration.
Measures of Intermediacy
If intermediate aphids owe their intermediacy to a change in em- bryonic development from that characteristic of one normal type to that of the other type (winged to wingless or wingless to winged), and if the differentiating structures are determined at different times, it must be expected that these structures would attain different grades of advancement in different individuals. These grades of advancement are, indeed, the criteria by which the time-of-determination theory must be judged. It is necessary, therefore, to have some means of recording their degrees of development. Objective methods of measuring struc- tures were employed when practicable. To the other characters nu- merical values were arbitrarily given, 0 representing the extreme found in one form, and 4 the opposite extreme in the other type. This classi- fication of the external characters was done on living animals which had been etherized. A binocular dissection microscope usually gave magnification enough for external examination, although there were times when a question arose concerning the presence of an ocellus pri-
440 KARL A. STILES
mordium, and this would be decided with the aid of a compound microscope.
\Yith respect to ocelli, 0 designates a complete absence of ocelli, which is typical for the wingless viviparous females; 4. the fully developed ocelli of a winged individual ; and values between these two extremes represent varying degrees of intermediacy. For example. 1 would mean ocelli one- fourth developed.
The light color of the antennae in the wingless female was rated 0. and the dark color of the winged as 4. with intergrades between these two numbers of 1. 2. and 3. An actual count of sensoria on each an- tenna was made; and while there was no effort to distinguish between right and left, each was tabulated separately.
Wings were used in this study as a standard of intermediacy for rough classification ; that is, reference to degrees of intermediacy without regard to internal structures was based on the amount of wing develop- ment. \Yhen an aphid is said to possess a low degree of intermediacy, it means that there is little wing development although the wing muscles might be- well developed; and conversely, when a high degree of inter- mediacy is designated, it means that the wing development is relatively u;reat irrespective of other structures concerned in intermediacy. The wing was chosen as a standard of reference because it is an external differential character which can most easily be rated. Wings were rated from 1 to .01, the former number meaning that an aphid was fully winged, and the latter that the wing was .01 of the normal length. Some wings were less than .01 of tin- length of normal wings; in fact, there was a fair-sized group in which the wings were represented only by a roughness of the thoracic region where wing buds should appear. These in the raw data were termed subintermediates, to indicate a very low degree of intermediacy. These subintermediates in Tables I to IX, are indicated by a negative sign (--) after .01, as .01 — .
The degree of internal intermediacy was judged by quantitative means. Km- this work transverse serial sections were- found most suit- able. Measurements of all the internal differential organs were taken with an ocular micrometer. In rating the ocelli, the diameter was first measured and then tin- number of sections counted.
The wing mu-culature in transverse sections (Fig. 2) consists of two inner and two outer vertical muscles, together with what appears to be a single rectangular mass of longitudinal muscles divided by a median fissure. Only the maximum width of the inner and outer ver- tical muscles was measured, but the total area of the longitudinal muscles was computed from measurements of both width and height. A record of the total number of sections containing vertical wing muscles was
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 441
made. The final quantitative rating of the wing muscle development was based largely on the width of the vertical muscles, for the correlation between width of muscle and number of sections was close enough to make the latter measurement of value largely as a check on the former. A careful comparison of the longitudinal and vertical muscles revealed that the longitudinal, proportionately, were consistently less developed than the vertical in intermediate forms. Because the correlation be- tween the two sets of muscles proved to be rather constant, little was to be gained by using the longitudinal muscles as a quantitative measure of intermediacy. That there might be a close correlation between these two groups of muscles seemed probable from a study of their develop- ment in the early embryogeny of the aphid, for it was learned that both appear at about the same time. At least, when the vertical muscles were found the longitudinal were always present, and both showed about the same relative development. This made it doubtful whether there was a difference in the time of determination for these two dif- ferent sets of muscles, although there could be.
The differential characters which proved of greatest worth as meas- ures of intermediacy in this investigation were the wings, ocelli, and wing muscles, and consequently these were most carefully rated. After the raw data on these structures had been obtained they were compared with the mean for the same organs in what were considered typical winged individuals. In tables of these values the development of these differentiating structures is expressed in percentages.
By means of the differentiating characters described above, a diag- nosis of the intermediates of this study is made with the results pre- sented in Tables I to IX. The data are grouped according to conclu- sions to be drawn, but for the present this may be ignored. The reasons for so arranging the data that conclusions are anticipated is that it saves repeating this material in a general table.
Expected Composition of Intermediates Under Various Suppositions as to Order of Determination and Direction of Change
As seen above, Goldschmidt has applied the time-of-determination theory to intersexes. In accordance with that theory, organs which are the first to be changed in successive offspring must be the latest to be determined in development, while the latest to change in successive offspring are the earliest to be determined. The order of embryonic segregation is the reverse of the order of modification in successive' offspring. If this theory is applicable to intermediate aphids produced
442 KARL A. STILES
during a transition from winged to wingless or wingless to winged, then it should be possible to determine the order of segregation if the direc- tion of change is known.
Whether the time-of -determination theory is correct in a given instance cannot be proven by finding that all characters of the two types of individuals change simultaneously. In fact, such a circum- stance would make possible the supposition that there was no difference in the time of determination of the different organs. But even tin nigh several structures did appear at the same time in the embryo, the pos- sibility remains that they may not all have been determined at the same moment, as there could be a difference in the intervals between the time of physiological determination of t\vo or more organs and their mor- phological appearance, and that difference might make possible their simultaneous appearance in development even though their determina- tions coincided in time. Nor is the time-of-determination theory con- firmed by the discovery that certain characters change in earlier offspring than do others. If intermediates that are near to the one extreme type always possess a given set of characters, and those near the other ex- treme type always have another set of characters, these facts furnish no support to the time-of-determination hypothesis. What is needed to prove the theory is to demonstrate ( 1 ) that when the intermediate con- dition is approached from one of the extremes, the changes in successive offspring shall take place in a definite order, and (2) that when the intermediate condition is approached from the opposite extreme the same order of change shall occur.
Whether the conditions necessary for confirmation of the time-of- determination theory are met in a given case of intermediacy can be tested by (1) considering all the possible types of individuals that would -ult from all conceivable orders of determination of the several dif- ferentiating structures concerned in intermediacy, and (2) tabulating the data on the quantitative development of each differentiating struc- ture in all the intermediates obtained. Then test which assumption each intermediate best fits. If more intermediates lit one assumption than others, that assumption is favored. With three different organs (wings, ocelli, and wim; muscles) to be considered there are six different pos- sible orders of embryonic determination which are shown below in the diagram as numbered assumptions. The order of determination is indicated by the arrous. that is, in 1 the wings are determined first, then wing muscle, and the ocelli last.
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 443
Six different possible orders of determination for three differentiating organs
Assumption I. Wings >• Wing Muscle — >• Ocelli
Assumption II. Ocelli — > Wing Muscle — >• Wings
Assumption III. Wing Muscle > Ocelli — >• Wings
Assumption IV. Wings > Ocelli — > Wing Muscle
Assumption V. Ocelli > Wings — > Wing Muscle
Assumption VI. Wing Muscle > Wings > Ocelli
For a given order of determination there will be a number of inter- grades as illustrated below in Assumptions IV A and IV B. The " A '' part of Assumption IV shows the various kinds of intermediate-winged aphids which would be produced as a result of a change from winged to wingless forms, with the order of determination wings, ocelli and wing muscle. Assumption IV B indicates the different kinds of inter- mediate-winged aphids produced with the same order of determination as in Assumption IV A, but during a transition from wingless to winged types. The third intermediate from the top of Assumption IV A would deviate from a normal winged individual by possessing normal wings, slightly reduced ocelli and much reduced wing muscle. If a detailed chart of each of the six assumptions is made out and then an attempt is made to determine which of the six possibilities for em- bryonic determination each intermediate best fits, it is obvious that each intermediate form will fit equally well two of the possibilities, depending on the order of determination, as will be illustrated later. If more intermediates fit one assumption than others, that assumption is favored. The above procedure was followed with the intermediate-winged aphids of this investigation. Three structures, wings, ocelli and wing muscles, showed enough intergradations in development to make them satis- factory for such a study. As previously stated, an effort was made to discover to which one of the six different possible orders of embryonic determination the intermediate-winged aphids of this research tabulated in Tables I to IX best conformed.
It will be seen that for each possible order of embryonic segregation there are two ways in which the characters may be combined, depending on the direction of change. The diagrams below illustrate the different combinations of intermediate characters as they occur in different winged-wingless aphids for selected assumptions. Each diagram showrs a complete transition from typical winged to wingless forms or vice versa. Read from left to right to determine the nature of any given intermediate.
If the transformation is from winged to wingless with the order of determination wings, ocelli and wing muscle, Assumption IV A, the
444
KARL A. STILES
wings would be most developed, the ocelli less, and the wing muscles least. If, however, the transition i> from wingless to winged with the order of determination the- same as in Assumption IV. then the wings will he least developed, tin- ocelli more, and the wing muscle most as shown in Assumption IV I'.. Thus it is possible to have two graduated series, one showing intermediates resulting from a change from winged
Assr.Mi'Tiox IIIA RKGARDIXG TIMK OF DETERMINATION Direction of change — winged to wingless
< 1'iubinations of intermediate chara<
Wing M :•
Normal
Normal
Noi mal
Slisjitlv reduced
Much reduced
\ ery much reduced
None
Ocelli
Normal
Normal
Slightly reduced
Much reduced
Very much reduced
None
None
Wings
Normal
Slightly reduced
Much reduced
Very much reduced
None
None
None
A>STMPTIOX IN1'> KKCAKDINC, TIMK OF DETERMINATION Direction of change — wingless to winged
Combinations of intermediate characters
\Ving Muscle
None
None
None
Slight Iv developed Mine .le\ eloped
I lighly developed
Normal
Ocelli
None
None
Slight \y developed
More developed
I Highly developed
Normal
Normal
Wings
None
Slightly developed
More developed
1 Highly developed
Normal
Normal
Normal
ASSUMPTION IVA I\K<;AKI>IX(; TIMF. OF DETERMINATION Direction of change — winged to wingless
'lunations of intermediate characters
Win
Noi m.il Normal
Normal
Slight l\
Mm li rediii nl
\ 61 • linn li reduced
V >ne
Ocelli
Normal
Normal
Slightly reduced
Murli i educed
Very much reduced
None
None
U'ing Muscle
Normal
Slightly reduced
Much reduced
Very much reduced
None
None
None
THEORY OF EMBRYONIC DETERMINATION IN APH1DS 445
ASSUMPTION IVB REGARDING TIMK OF DETERMINATION Direction of change — wingless to winged
Combinations of intermediate characters
Wings —> Ocelli — > Wing Muscle
None None None
None None Slightly developed
None Slightly developed More developed
Slightly developed More developed Highly developed
More developed Highly developed Normal
Highly developed Normal Normal
Normal Normal Normal
to wingless, and another including totally different kinds of intermediates when the change is from wingless to winged. The aphids of these two series, Assumptions IVA and IVB, are complementary in their com- position ; that is, in one the wings are large, in the other small, and the same relationship holds for the other structures. Figures 7 and 8 pic- ture this complementary condition in the wing muscles ; Fig. 7 is an intermediate with wings .60, and degenerate wing muscle .03 developed, while Fig. 8 is rated with wings .22 and non-degenerate wing muscle .66.
As indicated above, any intermediate aphid should fit equally well two of the six assumptions for different orders of determination because an intermediate of a structural composition that should result from an assumed order of determination, such as wing muscle, ocelli, wings, Assumption III B, could also result from the reverse order of deter- mination, wings, ocelli, wing muscle, Assumption IV A, if the direction of change were reversed. The above-mentioned assumptions illustrate this, for the composition of an aphid which fits Assumption III B is identical with one which fits Assumption IV A. In other words, an intermediate produced as a result of a transition from winged to wing- less in Assumption IV A is the same in composition as one resulting from a conversion from wingless to winged in Assumption III B. Hence, all that is shown by such a classification and testing of data would be that the order of determination was one of two possibilities, one the reverse of the other, and unless the direction of change is known the elimination of one of these possibilities is out of the question.
It was not known from the conditions under which these inter- mediate-winged aphids were reared whether a given intermediate was produced as a result of change from winged to wingless or the reverse. Unlike the gamic-parthenogenetic intermediates found by Shull, there was no gradual change from one type to the other. Shull found it possible to get stocks producing mostly gamic females and then effect
446 KARL A. STILES
a gradual change to parthenogenetic types with many intermediates ap- pearing during the change. Presumably the intermediates produced under these conditions were the result of a gamic-to-parthenogenetic change. In the experiments in \vhich intermediate-winged forms ap- peared there were usually both winged and wingless aphids with a few intermediates. Several females were placed on one plant and one aphid might be giving birth to offspring which were intermediate as the result of change in one direction while another could be producing inter- mediates a> a result of change in the opposite direction. And even if a female were to be reared in isolation on a plant, little could probably be told about the direction of change which resulted in intermediacy for the winged and wingless offspring occur in a very erratic manner through the family. It is also conceivable that in a given experiment an aphid could produce intermediate offspring as the result of a change in one direction at one time, and other intermediates from the opposite dirertinn of change at another time. If this occurred there would be no way by which the direction of change could be detected from the daughters. Shull (1928, 1929) has shown that light and temperature modify the development of wings but these agents cannot be relied on to change the course of development in all individuals in the strains now available which are producing the intermediates. Thus it will be seen that the experiments themselves afford no reliable method whereby the direction of change can be ascertained.
However, this investigation has revealed what is considered to he a means of deciding the direction of change which results in interme- diacy. As has been stated above, there were found to be two different classes of intermediates based on internal structures, namely. (1) those with histologically normal wing muscles and (2) those with degenerative wing muscles. It is postulated by the author that degenerate muscles are the coiiM-t|uencc of a transformation from winged to wingless forms. I J'-velopmeiit is assumed to have begun with factors favoring wing production dominant, then a physiological change taking place which nut MH]V inhibited further development of characters associated with wings, but brought about more or less degeneration of the wing muscle already produced. \Vhen a change takes place in the opposite direction. that i>, a conversion from wingless to winged, there is no degeneration of the organ- of the winged type because in this transition the physiology changes from one in which those organs are not present to one which favors their development. \Yith conditions favorable for the produc- tion of the strm -ture-> characteristic of winged females, these organs undergo normal histological development. The facts are that about 56 per cent of the intermediates of this investigation were characterized
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 447
by normal muscles, while about 44 per cent possessed degenerate muscles. This interpretation of the two classes of intermediates based on the nature of wing muscle affords a means of determining the direction of change in a given individual so that it is possible to classify each as the result of a conversion from winged to wingless or wingless to winged. In Tables I to IX inclusive, all aphids listed under the " A " assumptions possess degenerate wing muscles; all those listed under the " B " as- sumptions are wingless-to-winged transformations.
By a study of these Tables (I to IX inclusive) the conclusion that normal and degenerate wing muscles are correlated with some direction of change is deduced from the way in which the tabulated material fits the various assumptions. If degeneration of muscles wrere a purely fortuitous phenomenon, with one aphid as likely as another to be af- fected regardless of the direction of transformation producing inter- mediacy, every Assumption " A," of which there are six, should in- clude some intermediates, which according to the conditions of their classification would possess degenerate muscles. This expectation is not realized. Moreover, Assumptions III and IV are complementary as discussed above, and about 50 per cent of the aphids of this inves- tigation conform to Assumption IV B ; yet there is not an aphid to be found conforming to its complementary Assumption III A. If degen- eration of wing muscle were determined by mere chance, it would seem very probable that some of the aphids classified in Assumption IV B would have come under the influence causing muscle degeneration, thus changing their classification to Assumption III A. The converse of this statement is equally true; if the wing muscle in intermediates postulated to be changing from wingless to winged were decided only by chance, there should also be aphids which would fit Assumption III B ; but none were found. There seems no other way to explain why aphids with the intermediate characters of Assumption IV B, normal muscle, should not have possessed degenerate muscles which would fit them to Assumption III A except that normal muscle is associated with a cer- tain direction of change which is here postulated to be from wingless to winged. Thus it will be seen that the histological character of inter- mediate-winged aphids is correlated with direction of change. Whether it means a winged-to-wingless or wingless-to-winged transition depends on the correctness of the theory that degenerating muscles result from a change in physiology to which the immature wing muscles are sus- ceptible.
This method of establishing the direction of change which results in intermediacy makes it possible to eliminate one or the other of Assump- tions III and IV which are complementary as far as the development
448
KARL A. STILES
S I-IX
Tables I. II, III, IV, V, VI. VII. VIII and IX contain the data regarding the development of the differentiating organs of intermediacy of winged-wingless aphids. The data are arranged, in so far as possible, to fit the various assumptions made.
TABLE I
Composition of intermediate aphid fitting Assumption I A. Order of determina- tion: wings— wing mu-ele-»-ocelli. Change from winged to wingless. Degenerate wing muscle.
|
Slide No. |
Characters to which Assumption I A is fitted |
Number of sensoria |
Antennal color |
||
|
Wing length |
\Ving muscle development |
Ocellar development |
|||
|
1196c (7) |
.50 |
.49 |
.41 |
5-6 |
0 |
TABLE 1 1
Composition of intermediate aphid fitting Assumption IB. Order of determina- tion: wings— »-wing muscle— »-ocelli. Change from wingless to winged. Normal wing muscle.
|
Slide No. |
Characters to which Assumption I B is fitted |
Number of sensoria |
Antennal color |
||
|
Wing length |
Wing muscle development |
Ocellar development |
|||
|
11 70 c (2) |
.03 |
.28 |
.35 |
5-4 |
0 |
TABLE III
Composition of intermediate aphids fitting Assumption II A. Order of de- termination: ocelli— »-wing muscle— *-wings. Change from winged to wingless. I )e- • •rate wmu muscle.
|
Hi,l,. X,,. |
Characters to which Assumption II A is fitted |
Number of sensoria |
Antennal color |
||
|
' ' dlar development |
Wing muscle devi-!"i>in<-Mt |
Wing length |
|||
|
1 |
.35 .35 .23 .15 .11 .23 .22 .35 .35 .35 .17 |
.23 .16 .20 .12 .06 .08 .20 .16 .33 .20 .16 |
.15 .10 .10 .08 .05 .04 .02 .02 .02 .01 .01 |
4-8 6-7 6-8 5-5 5-6 6-5 7-7 10-5 10-5 4-4 5-5 |
0 0 0 0 0 0 (1 1 1 0 0 |
|
226 e |
|||||
|
\ |
|||||
|
.... <><><> ., |
|||||
|
4d . |
|||||
|
90S 1. 1 |
|||||
|
905 1. ; |
|||||
|
1016 a |
|||||
|
1087 |
|||||
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 449
TABLE IV
Composition of intermediate aphid fitting Assumption II B. Order of determi- nation: ocelli— »wing muscle— *-\vings. Change from wingless to winged. Normal wing muscle.
|
Slide No. |
Characters to which Assumption II B is fitted |
Number of sensoria |
\nicnnal color |
||
|
Ocellar development |
Wing muscle development |
Wing length |
|||
|
490 g (1) |
.66 |
.91 |
.95 |
3 |
|
TABLE V
Composition of intermediate aphids fitting Assumption IV A. Order of determi- nation: wings— >ocelli— »-wing muscle. Change from winged to wingless. Degenerate wing muscle.
|
Slide No. |
Characters to which Assumption IV A is fitted |
Number of sensoria |
Antennal color |
||
|
Wing length |
Ocellar development |
Wing muscle development |
|||
|
P-27 . . |
1.00 .80 .80 .80 .75 .75 .70 .70 .65 .60 .60 .60 .60 .60 .60 .50 .50 .50 .50 .40 .40 .40 .35 |
.95 .59 No sections of head .35 .50 .47 .47 .59 .59 .46 .46 .46 .46 .46 .41 .46 .35 .46 .46 .33 .33 .33 .23 |
.90 .12 .23 .08 .49 .12 .28 .45 .57 .08 .28 .06 .03 .41 .23 .41 .01- .25 .16 .08 .08 .16 .01- |
5-6 6-6 7-7 8-8 5-6 5-7 13-13 7-7 7-8 7-7 6-5 6-? 6-6 6-6 5-4 9-6 7-7 5-6 5-6 7-8 |
0 0 1 2 0 0 3 0 1 1 0 1 0 0 0 0 0 1 0 0 2 |
|
285 e (2) |
|||||
|
285 e (1) |
|||||
|
286 f |
|||||
|
282 e |
|||||
|
288 e (1) |
|||||
|
1196 c (2) |
|||||
|
428 d (2) |
|||||
|
235 b |
|||||
|
296 b (3) |
|||||
|
285 c (2) . |
|||||
|
279 e |
|||||
|
220 e |
|||||
|
1085 g (1) |
|||||
|
1196 c (11) |
|||||
|
488 d (2) |
|||||
|
220 d |
|||||
|
1021 a |
|||||
|
1188 b |
|||||
|
282 c (1) |
|||||
|
212 b . . . . |
|||||
|
1130f |
|||||
|
280 c (2) |
|||||
*JS%
&*•&
450
KARL A. STILES
TABLE VI
Composition of intermediate aphids fitting Assumption IV B. Order of determi- nation: wings— »-ocelli— »-wing muscle. Change from wingless to winged. Normal wing muscle.
|
Slide No. |
( li whic li Assumption IV B is fitted |
Number of sensoria |
Antennal color |
||
|
Wing length |
Ocellar development |
Wing muscle development |
|||
|
1 1 68 a |
.60 .60 .50 .50 .40 .40 .35 .30 .25 .25 .25 .22 .20 .20 .20 .12 .10 .10 .10 .09 .08 .05 .04 .04 .04 .03 .03 .03 .02 .02 .02 .02 .02 .02 .015 .01 .01 .01 .01 .01 .01 .in .01 .01 |
.71 .71 .94 .59 .46 .46 .46 .46 .46 .71 .35 .46 .35 .46 .46 .35 .71 .59 .59 .35 .35 .35 .23 .11 .11 .35 .23 .23 .35 .33 .23 .23 .11 .35 .23 .23 .23 .20 .23 .23 .28 .17 .23 .23 |
.83 .74 .99 .74 .90 .66 .83 .58 .49 .74 .45 .66 .63 .57 .62 .41 .99 .70 1.00 .66 .46 .49 .37 .23 .33 .42 .42 .42 .40 .62 .33 .25 .25 .49 .25 .33 .41 .33 .37 .37 .41 .37 .28 .28 |
5-8 14-15 11-11 9-8 6-6 6-6 7-7 5-6 5-8 6-6 5-6 5-5 12-9 7-7 11-7 7-8 6-8 6-6 6-6 6-7 4-4 8-8 7-5 6 or 7 7-8 6-7 5-5 6-7 5-5 6-6 5-5 5-7 4-7 5-7 |
0 0 4 3 0 1 0 0 1 0 0 1 0 0 0 0 3 3 0 0 1 0 0 0 0 0 1 1 1 0 0 0 1 1 0 0 0 0 0 0 0 0 |
|
472 c (1 ' |
|||||
|
1061 e (1) |
|||||
|
307 a (in |
|||||
|
472 e- > |
|||||
|
119(,r 4 |
|||||
|
597 <l (6 |
|||||
|
1196 c 1 |
|||||
|
44 c 1 |
|||||
|
1077 .1 |
|||||
|
905 1> '3) |
|||||
|
1IH>3 K (2) |
|||||
|
414 c |
|||||
|
472 c 1 |
|||||
|
1087 g (1) |
|||||
|
1174 c |
|||||
|
in<>7 f |
|||||
|
717 1, |
|||||
|
1015 f |
|||||
|
472 c i4> |
|||||
|
44 c (3) |
|||||
|
472 c (8) |
|||||
|
895 d |
|||||
|
267 d |
|||||
|
597 d 1 |
|||||
|
47 > < ' 14) |
|||||
|
589 d (3; . |
|||||
|
44 c M i |
|||||
|
44 c 1 |
|||||
|
1 1 |
|||||
|
597 ,1 (2) |
|||||
|
5'.') .1 <) .. . . |
|||||
|
295 c (2 |
|||||
|
•ni5 b (1) |
|||||
|
v,<> e 1 |
|||||
|
1111 a (1 |
|||||
|
KIM g (1) |
|||||
|
10.<5 .1 (9 |
|||||
|
1079 |
|||||
|
1079 . l |
|||||
|
Ui7') . > |
|||||
|
107') ; f, |
|||||
|
1079 |
|||||
|
I035d 1 |
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 451
TABLE VI (continued) Composition of intermediate aphids fitting Assumption IV B
|
Slide No. |
Characters to which Assumption IV B is fitted |
Number of sensoria |
Antennal color |
||
|
Wi ng length |
Ocellar development |
Wing muscle development |
|||
|
589 d (8) |
.01 .01 .01 .01 .01 .01 .01 .01 .01- .01- .01- .01- 0 0 0 0 0 |
.28 .23 .17 .23 .23 .23 .35 .23 .11 .23 .23 .11 .46 .59 .70 .82 .58 |
.37 .29 .29 .37 .37 .33 .49 .41 .37 .33 .33 .23 .58 .74 .72 .91 .66 |
7-6 7-7 8-7 6-6 5-5 5-5 4-4 5-4 3-5 4-6 8-5 13 8-9 12-13 13 |
0 0 1 0 0 0 0 0 0 0 0 0 4 2 3 4 |
|
589 d (7) |
|||||
|
589 d (5) . |
|||||
|
1103 d (1) |
|||||
|
1103d (3) 1103 d (4) |
|||||
|
1103 d (5) 1087 g (2) |
|||||
|
1085 d (8) |
|||||
|
1035 d (6) |
|||||
|
1035 d (7) . |
|||||
|
1035 d (5) |
|||||
|
307 a (7) |
|||||
|
773 b |
|||||
|
308 a (4) |
|||||
|
454 b |
|||||
|
307 a (3) |
|||||
TABLE VII
Composition of intermediate aphids fitting Assumption V A . Order of determina- tion: ocelli— *-wings— >-wing muscle. Change from winged to wingless. Degenerate wing muscle.
|
Slide No. |
Characters to which Assumption V A is fitted |
Number of sensoria |
Antennal color |
||
|
Ocellar development |
Wing length |
Wing muscle development |
|||
|
1017 c |
.94 .35 .35 .59 .23 .35 .23 .46 .35 .23 .23 .23 .11 .11 .35 |
.45 .30 .20 .20 .15 .15 .10 .10 .10 .06 .05 .05 .03 .02 .02 |
.33 .01- .01- .01- .01- .01- .08 .01- .01- .01- .03 .04 .01- .01- .01- |
5-9 6-7 6-7 6-6 5-7 5-7 5-5 6-6 6-6 6-6 6-7 6-7 6-5 |
0 1 0 1 0 0 0 0 0 0 2 0 |
|
224 c |
|||||
|
227d |
|||||
|
1162c |
|||||
|
44 c (6) |
|||||
|
287 e(l) |
|||||
|
44 c (5) . |
|||||
|
219c. |
|||||
|
228 e |
|||||
|
295 d (1) .... |
|||||
|
296 b (2) .... |
|||||
|
40 c (3) |
|||||
|
270e (1). . |
|||||
|
295 c (3) |
|||||
|
887 d... |
|||||
452
KARL A. STILES
TAMU-: \ III
Composition of intermediate aphids fitting Assumption V B. Order of determina- tion: ocelli— »wing:=— »-wing nui-cL'. Change from wingless to winged. Normal wing muscle.
|
Slide No. |
\\lin li Assumption V B is fitted |
Number of sensoria |
Antennal color |
||
|
liar development |
Wing length |
Wing muscle development |
|||
|
1091 d (1 ) |
.47 .47 .59 .33 .33 |
.70 .70 .60 .45 .40 |
.83 .86 .67 .58 .49 |
7-7 9-9 9-10 7-11 5-4 |
1 1 1 0 0 |
|
1091 d (2) |
|||||
|
589d (2) |
|||||
|
108'' |
|||||
|
597 d (5) |
|||||
TABLE IX
(.'omfwsition of intermediate aphids not fitting any time- of -determination assumption. Degenerate wing muscle.
|
Slide No. |
Characters to which no Assumption is fitted |
Number of sensoria |
Antennal color |
||
|
\Ving length |
Ocellar development |
\VinR musile development |
|||
|
'14 c |
.40 .01 - .01- .01- |
.40 .23 .11 .23 |
.33 .01- .01- .01 - |
5-6 |
0 0 |
|
490 g (3) |
|||||
|
I'-.W |
|||||
|
490 g (2) |
|||||
of tin- several organs is concerned. Kor example, aphids conforming to Assumption III \\ show the same- composition as those- conforming to Assumption IV A; they differ only in the direction of change as •tblishcd by the histolo^iVal character of the wing muscle. In As- sumption III I',, tin- change is from wingless to winged; in Assumption IV A it is from winged to wingless. If degenerate muscles are con- sidered the result of a change from winged to wingless, an individual with a given composition fitting equally well both Assumptions III B and IV A would be placed in IV A if the muscles were degenerate. There i> one outstanding piece of evidence that degenerate wing muscles do indicate a transition from winged to wingless aphids. Comparing the first few aphids tabulated in Tables Y and \ I, it will be observed that the wing development is what one would expect if the wings were determined first as here proposed. At the top of Table V (winged to amglc.ss ) is listed an intermediate with fullv developed wings, which is possible if \\ings are determined before wing muscles, but could not
THEORY OK EMBRYONIC DETERMINATION IN APHIDS 453
occur if muscles were determined before wings. On the other hand, the first aphid listed in Table VI (wingless to winged) shows only .6 of normal wing length. It is probably not without significance that the first few aphids listed in Table V show considerably more wing development than the first few listed in Table VI. If wings are deter- mined first, then in a winged-to-wingless transition the wings might be expected to show on the whole more development than in a wingless-to- winged change. The other possibility would be to postulate that wings are determined last, which does not fit the facts, as such an hypothesis would rule out the combination of fully developed wings and reduced wing muscle. More than that it is hardly conceivable even in an aphid with the composition shown by No. 285 e (2), Table V, wing length .8, ocellar development .59, and wing muscle development .12, that wing muscle could have been determined first, which the alternative theory would require. Thus all available evidence points to the conclusion that wings are determined before wing muscle, and that degenerate wing muscle characterizes a transformation from a winged to a wingless aphid in development.
So far in this paper no attempt has been made to prove that the ocelli are segregated at a time between the wings and wing muscles. Some evidence for this position of the ocelli in the order of determina- tion is available. To arrive at this evidence it will be helpful to show what would occur in the event that ocelli were determined between wing and wing muscles at an arbitrarily set time. In the chart below it is assumed that the order of determination is wings, ocelli, and wing muscle. There is also the assumption made that no overlapping in times of segregation for the several organs occurs. The times indicated are purely hypothetical as nothing is known about the actual time involved.
A.
Change from winged to wingless Wing determination Ocellar determination Wing muscle determination
1 :00 o'clock 2 :30 o'clock. Change . 3 :00 o'clock
to wingless Fully developed One-half developed No development
B.
Change from wingless to winged Wing determination Ocellar determination Wing muscle determination
1 :00 o'clock 2 :30 o'clock. Change 3 :00 o'clock
to winged No development One-half developed Fully developed
454 KARL A. STILES
The ocelli in .-] are only one-half developed because a change in physiology unfavorable to their development took place when they were only half segregated in the embryogeny. In B the ocelli are only one- half developed because a change in physiology which made their devel- opment possible in the embryogeny did not occur until the period of determination was half completed.
From this illustration it is evident that under the conditions postu- lated, the ocelli may not be very different in amount of development in two aphids derived from opposite directions of change, but the wings and wing muscles would show a considerable difference. If again Tables V and VI are compared it will be found that the facts conform to the theory. For example. Table V, slide 285 e (2), wing length .80, ocellar development .5^. wing muscle development .12, compared with Table VI, slide 307 a (10), wing length .50, ocellar development .59, wing muscle development .74. shows exactly the same relationship as illustrated above. Not all of the aphids fit the conditions as perfectly as the OIK-* drawn for illustration but the general trend is clearly the same.
Thus the facts support the theory that of the three organs, wings, ocelli, and wing muscle, the ocelli are median in the time of embryonic segregation.
1 1" the time-of -determination theory were perfectly applicable to in- termediate-winged aphids, and the intermediates were to be classified according to the assumptions of Tables I to IX inclusive, which are based on the time-of -determination theory and the direction of change established by the character of the muscle, then all intermediates would meet the conditions which parts "A" and ' B "' of one assumption entail. Examination of Tables I to IX. which are an attempt to dis- cover the applicability of the time-of-determination theory when the foregoing conditions are imposed, shows that the' material does not all tit any one assumption. However, Assumption IV is applicable to about <>> ]•<•]• cent of the aphids of this research. Other assumptions which are applicable to intermediates are I (2 per cent ), II (10 per cent), and V (lf> per cent). This would suggest that the time-of-determination theory has at least a limited applicability to the development of inter- mediate-winged aphids. In fact, the large percentage conforming to this scheme probably means that the assumption that the composition of intermediates i> determined by the times of embryonic segregation of their differential structures is in a general way correct. Were then' no Mich relation between the organs of intermediacy a much smaller pro- portion of intermediates should be of the composition shown in Tables \ and VI. As previously stated, there are 6 conceivable assumptions
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 455
for these intermediates. If the combinations were entered into at ran- dom only one-sixth of them should be of the kinds listed in Assumption IV (Tables V and VI). Their actual numbers are over four times this expectation.
Besides the 84 intermediates which conform to a single assumption regarding times of determination (Tables V and VI), there are 34 which fit other assumptions (Tables I, II, III, IV, VII, and VIII), and 4 that could not be classified under any one of these schemes (Table IX). It will be noted concerning these aphids not meeting the condi- tions of Assumption IV that most of them have been placed in another assumption due to the fact that they appear to show a shift in the time of determination of the ocelli. In most instances the development indicates a determination of ocelli before wings, while in two cases the amount of ocellar development indicates their determination after the wing muscles. This variation of the indicated time of determination of the ocelli in both directions suggests again that the usual time of segregation is between the wings and wing muscles.
The simplest explanation of the irregularities in the order of deter- mination found in Tables I, II, III, IV, VII, and VIII, after eliminating certain possible errors, is to postulate that there actually is variation in the time at which segregation takes place. The variation on at least a few occasions is so great that the time when the wings are determined is changed from the first of the three to the last. However, of particular interest is the fact that in all of the material studied not one case was found in which wing muscles were indicated to be determined first. It may be significant that there are far more individuals in which there was a change in the usual time of determination of only two structures as contrasted with those in which the position of all three was disturbed. For it would seem reasonable if this explanation is correct that slight irregularities would occur more frequently than those which involved more radical departures from the usual. The small group of four aphids (Table IX), in which two organs are of about the same development are exceptional individuals in which a change in the time of segregation of only one structure may have caused the irregularity. That the time of determination is not an unalterable fixed thing is shown not only by these irregular classifications, but in Tables V and VI, to which the intermediate group as a whole best conforms, there is considerable variation ariiong the individuals listed.
456 KARL A. STILES
AN INTERPRETATION OF THE COMBINATIONS OF CHARACTERS IN INTERMEDIATE Arums BASED ON THE PHYSI- OLOGICAL LEVEL THEORY
Inasmuch as the combinations of characters in the intermediate aphids of this report <1» not conform perfectly to the time-of-dctermina- tion theory, an analysis based on the physiological level theory would seem indicated. It will be remembered from an earlier statement that this theory postulates that the degree of intermediacy depends upon the state of sonic physiological process during development. It could de- pend on the quantity of some substance (possibly a hormone) present during the embryogeny. If all the differentiating structures reacted to exactly tin- same extent to a given quantity of this substance, or to the "level" of this physiological process, there would not appear to be a linear order of determination; all structures would give evidence of having been determined at the same time, or during the same part of the embryogeny. According to this theory without modification, when a given degree of intermediacy in one structure occurs, each other organ should show a definite degree of intermediacy. For example, if one intermediate at a given physiological level shows .1 wing, .2 ocellar, and .4 wing muscle development, every intermediate with .1 wing de- velopment should have the same grade of ocellar and wing muscle development. If, however, wings start development at a given physio- logical level, let us say when there is a certain concentration of a sub- stance, while ocelli and wing muscles do not, there arises a situation in which a difference in the time of determination of these organs is sur- mised, although such a difference may not exist. In other words, there may be a difference in the threshold of stimulation of the intermediate structures causing their appearance in different degrees. This explana- tion denies any importance to differences in the time of embryonic •ation of differentiating parts. There could be such differences even if there were a difference in the threshold of stimulation of the several organs, but it would not be the differences in time of determina- tion of the different parts that would determine the composition of intermediate individuals, rather it would be the differences in the thresh- old of stimulation of the differentiating parts.
A casual examination of the tables of intermediate-winged aphid data indicate- at once that the physiological level theory without modi- fications is not applicable to these aphids, for there is apparently no con- stant relationship between the degrees of development of the several structnn-s such as this theory requires. But if it is assumed that there is a dill- in the threshold of stimulation of the several organs in dif-
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 457
ferent individuals then there should be a difference in the development of a given structure in different intermediates. If Tables V and VI, which include nearly 70 per cent of the material of this investigation, are studied, it will be found, to take a typical example from Table V, slide 1196 c (11), that wings are .6, ocelli .41, wing muscle .23 in de- velopment. This would mean, according to the physiological level hy- pothesis, that the wings possessed a lower threshold of stimulation than either of the other two structures. Now if a comparison is made with Table VI, slide 1168 a, in which the wings are .6, ocelli .71, and wing muscle .83 developed, it will be seen that in this case the wing muscle would have to possess the lowest threshold of stimulation, ocelli the next lowest, and the wings the highest. It will be noted in the latter illustration there is required a reversal of the order of relative levels in the thresholds of stimulation of the several structures ; the threshold of stimulation of the wings in the second illustration is highest, whereas in the first illustration it was lowest. Thus, if this proposal of a varying threshold of stimulation is offered as a modification of the physiological level hypothesis in explanation of the combination of characters in the majority of intermediates in this work, it is necessary to postulate not only that the threshold of stimulation is different in different organs for the same individual, but that it is different for the same organ in different individuals, and that for most intermediates there is an alter- nating though definite order in which the threshold of stimulation varies. Other aphids which do not conform to Tables V and VI would necessi- tate the assumption of an order in which the thresholds of stimulation in the several organs vary. It is conceivable that the thresholds of stimulation could vary for the several structures in different aphids in such a manner as to produce a series fitting the time-of-determination scheme, but it seems highly improbable that at one time the threshold would so vary as to give individuals of a composition that would con- form to the winged-to-wingless direction of change and at another time individuals of such a composition that they would fit a wingless-to- winged direction of change. The difficulties involved in making this assumption provide the greatest objection to this hypothesis.
DISCUSSION
With the exception of the work of Shull on aphids, most of the previous investigations concerning intermediacy in insects have dealt with intersexes, that is, forms intermediate between the male and female of the species. The chief contributor in this field has been Goldschmidt (1931), who, in an analysis of intersexuality of the gipsy moth, Ly-
458 KARL A. STILES
inantria dispar. considers intersexuality in these moths as depending chiefly on general sex genes.
The intermediacy of aphids which has been a subject of considerable study at the University of Michigan is in sharp contrast in some re- spects to that of the ninth reported by Goldschmidt. In the first place, the type of intermediucy which we have studied in aphids is not an intersexual phenomena. t"»r we are dealing with only one sex. Secondly, it is highly improbable that it is due to a quantitative genie influence as postulated for Lyuuintria, for the aphids of this research were all parthenogenctic females, and in a parthenogenetic line there is a mini- mum of opportunity for genetical variation. Certainly the differences in individuals of a single strain of a parthenogenetic species are less likely to result from genetical causes than in an animal which repro- duces hiscxually as in the gipsy moth where different geographical race> were involved in the crosses. Inasmuch as it has been found pos- sible by modification of the environment to change the percentage of winged to wingless aphids, it would seem probable that intermediacy was due to some physiological state induced by environmental factors. However, Goldschmidt's assumption that embryonic determinations oc- etir at different times in development does with modifications seem ap- plicable to aphids, as has been discussed above. Modifications from a perfect application of the time-of -determination theory to aphids are necessary because the data from intermediate-winged aphids not only -bow considerable variation in the times of determination within the usual order of embryonic segregation but this sometimes is disturbed by irregularities.
Thus it will be seen that in two types of intermediacy which are quite different in that one is of an intersexual character and results from a genie imbalance, the other not intersexual and probably due to environmental factors, yet both support in a general way the theory that in at least some of the invertebrates there is a succession of em- bryonic determinations.
SUMMARY
In two strains of aphids of the species Macrosiphum solanifolii. in- termediates between winged and wingless females were experimentally produced, and a histological study was made of these. Aphids may show internal intermediacy when there is no evidence of it externally, there) ore the degree of winged-wingless intermediacy cannot be safely judged by external observations alone.
'I wo theories attempting to explain the combination of characters (wings, ocelli, \\-ing muscle) found in intermediate-winged aphids are
THEORY OF EMBRYONIC DETERMINATION IN APHIDS 459
discussed. The first theory postulates that the combination of charac- ters in intermediates is due to the fact that the differentiating organs are determined at different times in the emhryogeny (time-of -deter- mination theory), while the second theory assumes that the structural character of intermediates results from a given concentration of some substance (physiological level theory) in development.
It could not be ascertained from the conditions under which thoc aphids were reared what direction of change (winged to wingless or wingless to winged), assuming there was a change, produced a given winged-wingless intermediate, but evidence is given which indicates that the character of the wing muscle is correlated with some definite change in the direction of embryonic development. The evidence makes it appear probable that degenerate wing muscle means that the inter- mediate aphid is the result of some physiological change from a winged to wingless form during the embryogeny ; whereas the intermediate with normal wing muscle appears to be the result of a change from wingless to winged during its development.
If the time-of -determination theory were perfectly applicable to intermediate- winged aphids they should all fit one (the same one) of the six possible assumptions for the time at which the wings, ocelli and wing muscles may be determined in the embryogeny. More than two- thirds of the aphids of this study did fit one assumption (Assumption IV) which is evidence that in a general way the composition of inter- mediates is determined by the segregation of their differentiating struc- tures at different times in the embryogeny. This indicates that organs are determined in a serial order in aphids, as has been found true for certain vertebrates. The order of embryonic determination of the structures distinguishing winged from wingless is wings, ocelli, and wing muscles.
The small differences between the amount of development in wings, ocelli, and wing muscles found in many cases is taken to indicate that the time of determination of these organs is close together. However, that the period of embryonic segregation for these three structures is of considerable duration is suggested by the large number of intergrades.
Most of the intermediates which did not conform to Assumption IV appeared to owe their irregular combination of characters to a shift in the time of determination of the ocelli. The simplest explanation of these irregularities is to postulate that the time of segregation may vary for a given organ in different individuals.
An interpretation of the combination of differentiating characters of intermediate-winged aphids based on the physiological level theory would involve the necessity of assuming that the threshold of stimulation
460 KARL A. STILES
for the development of these organs varies in different individuals. Moreover, the threshold of stimulation would have to show a series of regular variations to produce aphids fitting one direction of development, as winged to wingless, at one time and another direction, as wingless to winged, at another time. The difficulties involved in making such an assumption renders this hypothesis highly improbable.
The results of this investigation support the time-of-determinatimi theory with the modification of frequent variations in the time at which an organ may he embryonically determined.
LITERATURE CITED
BAKFR, A. C. ANH \Y. F. TURNER, 1916. Some intermediates in the Aphididse
(II. i ). Proc. Ent. Soc. Wash., 18: 10.
BARDEEX, ('. R., 1900. The development of the musculature of the body wall in
tin- pig including its histogenesis and its relations to the myotomes and
to tin- ^kek'tal and nervous apparatus. Johns Hopkins Hasp. Rep., 9: 367. Gou'.M H \iinr, R., 1°31. Analysis of inti-rsrxuality in the gipsy-moth. Quart.
Rev. Biol, 6: 125. JAM r, ("HAKI.KS, 1907. Histolyse, sans phagocytose, des muscles vibrateurs du vol.
clu-x les reines des Fourmis. Comfit. Rend. Acad. Sci., Paris, 144: 393. MANi.iii.it, OTTO, 1928. Probleme der Entwicklungsmechnik. Natunviss., 16: 661. MOTTRAM, J. C., 1913. On the action of beta and gamma rays of radium on the
cell in different states of nuclear division. Arch. Middlesex Hasp., 30: 98. PA i I INI M., 1915. Pink and green aphid of potato. Maine Agr. Exp. Sta.,
University of Maine, Bull., 242: 205. SHULI., A. F., 1928. Duration of light and the wings of the aphid Macrosiphum
solanifolii. Arch. Enlzv.-mcch., 113: 210. SHUI.I., A. F., 1929. The effect of intensity and duration of light and of duration
of darkness, partly modified by temperature, upon wing-production in
aphids. Arch. Entiv.-mecli., 115: 825. SIIUI.L, A. F., 1930. Order of embryonic determination of the differential features
of Liamic and partheno^em-tic aphids. Zcitschr. ind. Abst. u. Vererb., 57:
92. SHULL, A. F., 1932. Clonal differences and clonal changes in the aphid Macro-
Mphum solanfolii. Am. Nat.. 66: 385. SHULL, A. F., 1933. The time of embryonic segregation in aphids as determined
from intermediate types. Proc. Nat. Acad. Sci., 19: 164. SHUI.I., A. 1'., 1937. The production of intermediate-winged aphids with special
ri-i'. < to thi problem of embryonic determination. Biol. Bull., 72: 259. II.. l''-'7. Xt IK- Arbeiten iiber Organisatoren in der tierischen Ent-
wicklung. Natttrwiss., 15: 946. SIII.KS, KARL A., 1934. Normal butyl alcohol tcchnic for animal tissues with
ice to insects. Stain Technology, 9: 97.
I i' HANCO, I I'.., 1''24. Studies on the embryology and postnatal development the Aphidid.-r, with special reference to the history of the "symbiotic .11," or myci-tonii-. Philippine Jour. Sci., 24: 143.
PHYSIOLOGY OF REPRODUCTION OF OSTREA VIRGINICA I. SPAWNING REACTIONS OF THE FEMALE AND MALE
PAUL S. GALTSOFF
(I' nun the U. S. Bureau of Fisheries, Published by Permission of the U. S. Commissioner of Fisheries)
Sexual cycles of many marine organisms are determined by specific rhythms in the development of their gonacls, and by environmental con- ditions which to a great extent control the discharge of ripe sex cells. The simplest sexual reactions are to be found among the animals which, lacking organs of copulation, shed their eggs and sperm directly into the water. In these cases fertilization is a matter of the chance meeting of ovum and spermatozoan outside of the organism, and the propaga- tion of the species is dependent upon the production of large numbers of sex cells and the correct timing of their emission. The latter condi- tion is significant for it enhances the probability that the eggs will be inseminated before they become unfertilizable. Since our knowledge of the physiology of the reproduction of marine invertebrates is very inadequate, conditions which govern the discharge of their sex products are but little understood.
The present investigation of the physiological reactions involved in the reproduction of the oyster was undertaken with the view of deter- mining the factors controlling the sexual activity of this mollusk. Be- cause of its widespread distribution, its abundance and the anatomical simplicity of its organs of reproduction, the common oyster of com- merce appears to be very suitable for such a study. The work was car- ried on during a number of summers at Woods Hole, Massachusetts ; Milford, Connecticut ; and Galveston, Texas. Supplementary observa- tions on O. gigas (Japanese oyster) and O. commercialis (Australian oyster) were made at the Hopkins Marine Station, Pacific Grove, California, and at the Marine Laboratory of the University of Hawaii in Honolulu. The first part of this study contains a description of the sexual reactions of the female and the male oyster. The subsequent parts, which will be published separately, deal with the stimulation and specificity of sexual reactions and their significance for the survival of the species.
REPRODUCTIVE CYCLE
The annual reproductive cycle in Ostrca viryimca begins in autumn immediately after the completion of the preceding spawning period. At
461
462 PAUL S. GALTSOFF
this stage the gonad follicles are lined with indifferent germinal epithe- lium and the sexes arc often indistinguishable. A transition to the next, the sex-differentiation stage, is marked by a slight expansion of the gonad tissue and the appearance of either ovo- or spermocytes. The gonad remains in this condition until the next spring when an extensive proliferation of the follicles accompanied by rapid ovo- or spermato- genesis results in the production of large numbers of sex cells. During the following summer almost all of them are discharged and the few remaining within the follicles are cytolized and destroyed by phagocytes.
The spawning season, during which one or several emissions may occur, lasts for about six weeks in the North Atlantic States (north of the Chesapeake P>ay) but in the warm waters of the South Atlantic and Gulf of Mexico may extend from March to the end of October. In every body of water a great part of the oyster population spawns more or less simultaneously. This is evidenced by the appearance in plankton collections of a large number of oyster larvae which often occur at regular intervals or cycles corresponding to the outbursts of sexual activity (Prytherch, 1928; Hopkins, 1931).
Young oysters, excepting those living north of Cape Cod, reach sexual maturity at the end of their first year (Coe, 1936) and barring unfavorable conditions, the mollusk breeds every year until its death. The fecundity of an oyster apparently does not decrease with age for in very large specimens measuring from 6 to 10 inches in length, the gonads may be well developed and contain large numbers of fertilizable eggs.
ANATOMY OF THE ORGANS OF REPRODUCTION
A brief description of the essential anatomical features of the gonads is necessary for an understanding of the mechanism of ovulation and ejaculation. In both sexes the gonads are paired structures consisting of a large number of profusely branching tubules situated within the layer of connective tissue immediately beneath the epithelial covering of the body. In a well-developed organ the ramifying tubules concresce along the dorsal side, forming one continuous layer investing the vis- ceral mass. Two distinct systems of genital canals and two gonoducts, one on each side of the body (Fig. 1, GD) remain the only signs of the paired condition of the gonad. The thickness of the gonad layer, mea-uivl in transverse section through the anterior part of the body at the level of the stomach, varies from year to year according to the number of o-lK produced during the winter and spring. In cases of extraordinarily prolific development this creamy and well-defined zone may reach over Jd mm. in thickness in an adult oyster. It may be less
SPAWNING REACTIONS OF OSTREA VIRGIN ICA
463
than 1 nun. in a poorly developed gonad. It is not clear what conditions are responsible for these differences, which may occur from year to year in the same locality and in the same stock of oysters. This problem calls for special investigation which is beyond the scope of the present
work.
In a well-developed gonad of an oyster, the branching and twisted tubules are crowded with sex cells. A few follicles can be found, the
w
FIG. 1. Female Ostrca vlrg'mlca, right shell and part of the mantle fold re- moved. The wall of the suprabranchial chamber dissected and gills pulled down to show the ovary and the water tubes of the gill lamellae. X 1.
A, anus; B, border of mantle; C, cloaca; E, eggs passed through the gills and accumulated in the pallial cavity; G, gills; GC, genital canals; GD, gonoduct (ovi- duct) ; H, hinge; M, mantle; O, ovary; P, pallial cavity; SB, suprabranchial chamber ; SM , adductor muscle, portion with striated fibres ; T, tentacles ; UM, adductor muscle, portion with smooth fibres; W, water tubes of the gills; UGC, urinogenital cleft. Arrows indicate the direction of the cloacal current.
lining of which still consists of a single layer of undifferentiated ger- minal epithelium or of cells at various stages of spermato- and ovo- genesis. The wall of each follicle is made of a thin layer of connective tissue fibres. The outermost portion of the gonad, adjacent to the surface epithelium, is occupied by an anastomosing system of genital canals which in a ripe specimen are easily recognizable with the naked eye. They form an arborescent structure (Fig. 1, GC) in which the branches increase in diameter toward the posterior end of the gonad.
464
PAUL S. GALTSOFF
The genital canals persist throughout the year and constitute the primary gonacls, which, by the growth of their inner walls and a series of cvaginations, give rise to the tubular secondary follicles. In transection the lumen of the genital canal appears as a narrow space between a thin body wall, consisting of a layer of connective tissue covered with epi- thelium, and a dense mass of ovarian or tcsticular follicles. The por- tion of the lumen adjacent to the body wall is lined with a ciliated epithelium while its npp.isite side, i.e., the inner arch of the lumen, is
CE
^*k \
M
•-.'--• - |*^vJV3g&,
. r* : tf -•'••• K
Vic,. 2. Ostrca cdidis. transxersc section of the right urinogenital cleft and oviduct preserved during o\ illation. Camera lucida drawing X 150. Hsematein- eosin.
/.'. hlood cells; CE, ciliated epithelium; CT, connective tissue; F, connective ti-Mic lihres; (,'. oviduct; M, muscle fibres; MC, mucous cells; R, ridge of ciliated liarating the oviduct from the urinogenital cleft; SB, suprabranchial cham- (), ripe eggs discharged from ovary; U, urinogenital cleft.
lined with germinal epithelium. Scattered muscle fibres are found here and there in the cnmiectivc tissue of the gonad but are slightly more abundant in the body wall exterior to the genital canals. In both sexes the canaK cnnvergc on each side of the gonad into a very short gonoduct (oviduct or >]>ermidurt) through which the sperm or ova are discharged into a urino<r.-nital cleft (UGC). In an adult oyster about 4 inches long, examine. 1 bv the author, the oviducts did not exceed 0.15 mm. in length while the spermiducts were about 0.75 mm. long. When the
SPAWNING REACTIONS OF OSTREA VIRGINICA
465
gonoduct is empty its lumen has a slit-like appearance with the ciliated lining folded in numerous ruffles (Fig. 2). In the ducts filled with eggs or sperm, the walls are greatly distended and the epithelial lining is stretched (Fig. 3).
The lumen of the gonoducts is completely lined with ciliated col- lumnar epithelium which rests on a hasal membrane surrounded by a layer of connective tissue fibres. The absence of germinal epithelium distinguishes the gonoducts from the genital canals although the transi- tion from the canal to the duct is gradual. Circular and oblique muscles
CT
MC
<SS&slll$fe&tK3&P's£i$SS&& -J^'o • ..!~
CE
FIG. 3. Ostrca cdiilis, transverse section of left urinogenital cleft and spermi- duct preserved during- ejaculation. Both spermiduct (G) and the urinogenital cleft ([/) are filled with ripe sperm. Camera lucida drawing X 150. Hsematein- eosin.
CE, ciliated epithelium; CT, connective tissue; G, spermiduct; MC, mucous cells ; M, muscle fibres ; SB, suprabranchial chamber ; U, urinogenital cleft.
are scattered in the walls but do not form a continuous sheathing. In preparations stained with hematoxylin and eosin, the connective tissue fibres can easily be mistaken for muscle fibres but their true nature is clearly revealed in sections treated with the differential triple stain of Mallory which stains the connective tissue blue. Roughley (1933) states that in Ostrca commercialis the gonoducts are surrounded by bands of sphincter-muscle fibres immediately internal to the cleft. No such structures have been found in 0. virginica although, as stated above, a few scattered muscles occur in the walls of the gonoducts.
466 PAUL S. GALTSOFF
Unfortunately no illustration of the sphincter is given by Roughlcy and a detailed comparison between the gonads of the two species of oysters is therefore impossible.
The urino-genital cleft into which the gonoducts open anteriorly and independently of the ureter is a narrow slit on the surface of the gonad (Fig. 1. I'd'C). A delicate- membrane of connective tissue covered on both sides with ciliated epithelium forms the external portion of its wall. Histologically the cleft differs from the gonoduct only in the character of its epithelial lining which contains large numbers of mucous cells absent in the gonoduct (Figs. 2, 3, MC, CE). In the female the opening of the g. .noduct into the cleft is marked with a well-pronounced ridge of tall ciliated cells not noticeable in the male (Fig. 2, 7?). On each side of the oyster the urinogenital cleft opens into a suprabranchial chamber in which a strong outgoing stream of water is maintained by the ciliary motion of the gill epithelium.
From an anatomical and histological study of the gonad it may be inferred that eggs and sperm are discharged primarily by the motion of the cilia of the epithelial cells lining the genital canals and gonoducts and that the contraction of the muscle fibres scattered between the follicles and in the walls of the canals is of lesser importance and prob- ably only facilitates their release from the follicles. Judging by the distribution of the muscle and connective tissue fibres in the walls of the gonoducts, it appears possible that their contraction may to a certain extent constrict the lumen and thus impede the emission of sperm or ova. The muscle fibres appear to be better developed in the walls of the spermiducts than in the oviducts (Figs. 2. 3, 37). Their possible role in ejaculation is discussed later.
SI-AWNINC RKACTIOXS OF TIIK FKMAI.K
The .spawning of the female may be divided into three distinct phases: the discharge of eggs from the ovary (ovulation), the rhythmi- cal contractions of the adductor, and the closing and opening of the mantle. Of these the first phase is obviously the principal sexual reac- tion of the organism, while phases two and three may be considered as accessory reactions providing a mechanism for the dispersal of eggs. Spawning may be observed in a ripe female without employing special apparatus or procedures. For observation the oyster is placed in a dish filled with warm sea water (25-27° C.). When the shell opens and the inollusk begins to feed, one can easily notice that the opposite borders of the mantle arc kept sufficiently apart to permit free a • nit,, the pal Hal chamber. Under ordinary circum-
SPAWNING REACTIONS OF OSTRKA VIRGINICA 467
stances this position of the mantle is maintained for a long time with only occasional changes (Hopkins, 1933, p. 482, Fig. 7, A). An ap- proaching spawning is marked, however, by an unusual muscular ac- tivity along the edges of the mantle and by a change in the position of its inner borders and tentacles. The opposite folds which previously were flattened and kept almost parallel to the surface of the shell, now occupy positions perpendicular to it. Contractions of the mantle and the corresponding changes in the positions of its borders begin simul- taneously from the posterior and anterior ends and spread toward the middle. The process continues for several minutes. During this time the borders may come in contact with one another and separate again. Finally they assume such a position that the pallial cavity is almost completely closed, only a small opening being left at a place approxi- mately equidistant from the mouth and cloaca.
In the majority of cases the tonus level of the adductor muscle changes during this time and the valves open more widely. Then the adductor begins a series of contractions which continue until the end of spawning. Coincident with the relaxations of the adductor and gaping of the valves, a white cloud of ova appears at the anterior part of the pallial chamber (Fig. 1, E} and, descending posteriorly, fills it. The ova accumulate in front of the small opening left between the two opposite borders of the mantle and by sharp contraction of the adductor are expelled into the water (Fig. 6). Almost immediately the adductor relaxes, the valves open again, the borders of the mantle come together leaving the opening at approximately the same place it occupied before, and a new batch of eggs, which in the meantime has entered the pallial chamber, is expelled. The process repeats itself with great regularity and may continue for more than one hour, its duration depending on the amount of ripe ova in the gonad. Gradually the discharge of eggs ceases; the contractions of the adductor become weaker and irregular, and the reaction reaches an end.
The activity of the mantle during spawning is less important than the contraction of the adductor for in many instances, especially when only a small number of eggs are discharged, the mantle borders remain flattened and the pallial cavity is open along its entire length. It is evident that the formation of a small aperture, although helpful in producing better dispersal of eggs, is not indispensable to the success of spawning.
For a more detailed study of the behavior of the adductor muscle during spawning the following technique was used. An oyster im- mobolized on its left side with a mixture of plaster of Paris and cement was placed on a suitable support in a glass tank containing 27.5 liters
468
PAUL S. GALTSOFF
of sea water. The right valve was attached to a lever which recorded the shell movement on a slow-motion kymograph. Temperature was maintained constant within ±0.1° C. by means of a heater, stirrer, and mercury temperature controller, operated through a relay. The entire set was mounted on an adjustable base supported by a tripod. Water was continually aerated or. if necessary, was kept running. The lever was arranged in such a way that the closed positions of the shell cor- responded to the peaks of the kymograph curve. By this method sev- eral hundred records of shell movement were obtained. Those repro- duced in this paper are the photographic prints of the originals.
B
LATENT PERIOD
O VU L ATION
I-ii,. 4. Ostrca virginica, spawning reaction of a female. A, normal muscular liehavior during feeding; B and C, spawning reaction. Time interval, 1 minute (bottom line). Xute the change in the tontis level (luring the latent period, cf indicates time when live sperm was added to the water.
The closing of the shell of an oyster is affected by the contraction of the adductor muscle; its opening is due to the elasticity of the hinge ligament which acts as a spring forcing the valves apart when the adductor relaxes. Spontaneous shell movements are therefore entirely eniitrolled by the contractions and relaxations of the adductor, whereas the purely mechanical action of the hinge depends exclusively on the expansion of the compressed elastic material of which it is composed. Under ordinary conditions the shell movements of an oyster are rather irregular ( Fig. I. ./ (. Occasional strong contractions of the adductor resulting in a complete closure of the valve are followed by periods of relaxation interrupted by one or several light contractions causing par-
SPAWNING REACTIONS OF OSTREA VIRGINICA 469
tial closing of the shell. Strong contractions are either ejection reac- tions by which the oyster gets rid of the material accumulated in the branchial chamber, or are responses to a great variety of chemical and physical stimuli ; as for instance — mechanical disturbances, sudden change in the intensity of illumination, chemical irritation by acids, salts, and various other organic and inorganic compounds. The mag- nitude of the reaction in these cases is usually directly proportional to the intensity of the stimulus.
Partial closure of the valves is caused by the contraction of the stri- ated part of the adductor muscle, while the maintenance of the valves in a given position against the pull of the hinge is attributed to the smooth component or so-called " catch " muscle. Under normal en- vironmental conditions shell movements are characterized by long re- laxation periods which vary from two to several minutes and are often interrupted by secondary contractions. A comparison of the records obtained during feeding with those performed at the time of spawning, reveals a great difference in the muscular behavior. The approach of ovulation is very often accompanied by a change in the tonus level (Fig. 4, B} and greater relaxation of the adductor. Then begins a series of rapid contractions and relaxations following one after another with remarkable regularity and continuing for a considerable time. Con- stancy in the amplitude of the relaxation curve, especially during the first half of the reaction, and the remarkable rhythmicity of the con- tractions are the most distinctive features -of the sexual reaction of the female. These phenomena do not occur under any other circumstances. Attempts to reproduce this type of reaction in non-spawning females by physical stimulation (pricking, electric current) or by applying various drugs and chemicals, have been unsuccessful. Rhythmical contractions of the striated component of the adductor in O. gigas were described by Hopkins (1936, p. 500, Fig. 1), who noticed their relations to the discharge of water from the cloaca and suggested that the impulses of the activity originate within the digestive system. That the rhythmicity of muscular behavior can be due also to other causes is obvious from the observations on oysters left for several hours without aeration or changing water (Fig. 5). In these cases a peculiar behavior of the adductor is probably caused by the accumulation of CO2 and other products of metabolism for the normal activity is restored as soon as water is renewed. Injection of 1 : 10,000 solution of adrenalin into the pallial cavity also results in the rhythmical contraction which, how- ever, lasts only for a short time (Fig. 5, C}. All these contractions have longer relaxation periods and are less regular than those associated
470
PAUL S. GALTSOFF
with the sexual activity of the oyster. This can easily he noticed by comparing the records reproduced in Figs. 4 and 5.
The question may he raised whether the behavior of the adductor during spawning represents ejection reactions in response to irritation caused by the presence of ova in the pallial chamber. That this is not the case is indicated by the continuation of typical spawning reactions after the discharge of ova has ceased. Furthermore, the quick relaxa- tion, which always precedes a sharp contraction of the adductor during spawning, begins before the eggs appear in the pallial chamber and the amplitude of the contraction is entirely independent of the amount of the discharged material. It is typical of the spawning reactions that
B
FIG. 5. Non-spawning rhythmical contractions in female oysters. O. vir- yinica (A) and O. gigas (B) kept for several hours without aeration. C, effect on adductor muscle produced by 1 : 10,000 adrenalin solution injected into pallial cavity of O. riryinica. Time of two treatments indicated by arrows. Time in- terval, 1 minute (bottom line).
the contractions begin with the maximum amplitude of the up and down strokes (Fig. 4), although at the verv beginning only a few ova are discharged. If the rhythmical contractions were only ejection reactions comparable to the oyster's response to irritation caused by the accumula- tion of foreign material in its pallial chamber, one would expect that the amplitudes of the curves would have varied with the amount of material expelled. All the records obtained during the present investi- gation show, however, that the strokes of the greatest amplitude occur at the beginning of the reaction and remain constant for a considerable period of time. The supposition that the contractions of the adductor during spawning arc ejection reactions is therefore untenable. It may be inferred that the adductor of the ovulating female receives specific
SPAWNING REACTIONS OF OSTRKA YlRdlXICA
471
internal stimuli which control its action, and which are present only during ovulation.
At the beginning of spawning the adductor muscle may contract as many as three times a minute hut gradually the rhythm slows down, the contractions become less frequent and less regular; the duration of the relaxation periods increases, and the muscle resumes its normal activity. Sometimes the change takes place so gradually that the end of the reac- tion can not accurately he determined.
FIG. 6. Female O. vir</inica in the act of spawning. The photograph was taken at the moment the eggs were discharged by snapping of the valves. Note near the edges of the shell, a narrow jet of eggs expelled through a small opening between the borders of the mantle.
Kymograph records show that during spawning the relaxation pe- riods are interrupted by brief pauses which give the curve a ladder-like appearance usually more pronounced toward the end of the spawning (Fig. 4, C). As will be shown later, this second half of each relaxation period coincides with the appearance of ova in the pallial chamber.
Contractions of the adductor, plus the action of the mantle, provide an efficient mechanism for the dispersal of ova in the surrounding water. The small hole between the two borders of the mantle through which
472
PAUL S. GALTSOFF
the nva an.- ejected. produces tlie same effect as a narrow nox/le at the end of a garden hose, considerably increasing the zone of their dispersal. Spawning females taken out of the water were observed to shoot their spawn to a distance of about four feet. Since fertilization takes place outside the organism and since the ova, being heavier than water, quickly settle on the bottom, the biological significance of this arrangement and the advantages it provides for the propagation of the species are obvious.
Fie. 7. Transverse section of the gills of the female O. rirginicn killed during spawning. Photomicrograph X 88. Haematein-eosin. Note the presence of eggs in the water tube. IT, and in the o-tiuin. O.
During (ivulatioii ripe eggs are discharged through tin- oviduct and urinogenital cleft into tin- suprabranchial chamber and cloaca. Their passage from the ovary and through the ducts can easily be observed in a livin- organism. For this purpose a spawning female is taken out of the water, one of its valves is removed, and tin1 urinogenital cleft ex- posed by niakiii-j an incision in the wall ol" the cloaca and suprabranchial chamber. The oyster is immediately returned to the water. Tin- pas- sage of eggs along the genital canals, oviduct ami tlirmi-li the cleft can now easily be observed with a magnifying lens. < >va expelled bv the
SPAWNING REACTIONS OF OSTRKA VIRGINICA
ciliary action of the epithelial lining of the ducts form a continuous stream and settle in the exposed part of the cloaca. Some of them are carried away by the cloacal current. Ovulation, once started, continues without interruption for a long time and is not stopped even by severe injury to the organism. The contractions of the severed adductor muscle persist but are not correlated with the delivery of eggs from the ovary. There are no noticeable contractions in the walls of the genital ducts or in the urinogenital cleft.
The question naturally arises as to how, in the intact organism, the ova discharged into the suprabranchial chamber reach the pallial chamber from which they are completely separated by the gills and walls of the branchial chambers (Fig. 1). Investigations of Kellogg (1892) ; Yonge (1926) ; Galtsoff (1928) ; Hopkins (1933) and others show that water is pumped from the pallial chamber through the ostia and water tubes of the gills into the suprabranchial chamber from whence it is expelled into the cloaca. In 0. virginica a portion of it goes into an asymmetrical promyal chamber. In an actively feeding oyster a stream of water can easily be noticed at the dorse-posterior and dorso-anterior sides of the body. One would naturally expect that every particle introduced into the suprabranchial chamber would be carried away with one of these streams (Fig. 1). Yet the ova of a spawning female find their way into the pallial chamber against the current produced by the gills and are expelled from the posterior side of the body (Fig. 6).
Observations on ovulation in females in which the valve has been removed and the cloacal wall has been dissected, show that the only way the eggs discharged into the suprabranchial chamber can reach the pal- lial cavity is through the water tubes and ostia of the gills. This conclu- sion is fully corroborated by examination of sections of female oysters killed during the act of spawning in which unfertilized ova were found inside the water tubes and near the ostia (Fig. 7, W, O). Since the beating of the lateral cilia throws the suspended particles toward the outer surface of the gill lamellae (Galtsoff, 1928) where they are caught by the frontal cilia and eventually conveyed toward the mouth, the pres- ence of an egg just underneath the lateral cilia is good evidence that it has reached this place from inside through the water tubes and supra- branchial chamber. This conclusion is fully supported by experiments in which the passage of eggs through the gills was observed in vivo. The following technique has been developed for making such observa- tions. A piece of shell of a ripe female was sawed off and carefully removed avoiding any injury to the tissues. After the operation, the underlying piece of mantle usually rolled up exposing the gills (Fig. 8, A}. The oyster was then placed in a finger-bowl filled with sea water
474
PAUL S. GALTSOFF
(25° C.) and ovulution was stimulated by the addition of sperm (Galt- soff, 1930/')- The shell movement was recorded on a kymograph by means of a lever attached to the remaining portion of the valve. The exposed gills were illuminated by a strong light and examined under a binocular microscope mounted on a suitable stand. Under these condi- tions the passage of eggs through the gills could easily be observed. It has been noticed that they emerge from the ostia in the area which
B
FIG. 8. Oslrca rirtjinicu, passage of eggs through the gills. ./. portion of is sa\\cd nlT and the mantle rolled up. I'Xgs (£) pass through the anterior part of the uilN. Ii , anterior half "f the shell is removed but the adductor is intact. Kggs are discharged through the cloaca.
- from tin- anterior part of the gills for about two-thirds of their length. The greater number of eggs pass along the proximal half of the gill lamella-, although some of them escape through the distal portion of tin- gill. Tliis experiment has been repeated several times and each time the moment of the appearance of eggs on the surface of the gills was marked on the kymograph curve. The emergence of eggs through the ostia coincided with the maximal gaping of the valves and was therefore associated with the rhvthmical contractions of the adductor.
SPAWNING REACTIONS OF OSTREA VIRGINICA 475
What forces the eggs to take the course through the narrow passages of the water tubes against the current produced by the gill epithelium? Several possible factors have been considered : a reversal of ciliary mo- tion in the gills, suction caused by the rhythmical movements of the valves, and difference in pressure between the suprabranchial and pallial chambers. While the evidence is not sufficiently conclusive to permit a definite answer to the question, certain of these possibilities may be ruled out.
The possibility of the reversal of ciliary motion cannot be verified by observation because the action of the lateral cilia cannot be observed in the intact gill lamella. The beating of the frontal cilia may, however, be observed by noting the movement of particles of carmine powder. During the passage of eggs it appears to be normal. If there is a reversal of ciliary motion it is therefore limited to the lateral cilia. The possibility of a reversal of the lateral cilia may not be completely ex- cluded, but confirmatory evidence is lacking. Nelson (1936) states that the rate of pumping of water greatly decreases during the spawning of the female oyster. This may be indicative of a disturbance in the functioning of the lateral cilia. Another explanation is possible, how- ever, for we know that the mantles of the spawning female reduce the opening, or even almost close the pallial aperture. This action obvi- ously will decrease or completely stop the exchange of water. The reduced rate of pumping during spawning does not, therefore, mean that the beating of the lateral cilia ceased or was reversed. Other ob- servations cast further doubt on the reversal of ciliary motion in the gills. Experiments were performed in which the anterior portion of the shell was cut away, preventing shell movement but leaving the muscle with adjacent part of the shell intact (Fig. 8, B). Because the hinge was severed, the oyster was unable to open its valves, but the adductor muscle and its nerve ganglia were in no way affected by the operation. Under these conditions eggs discharged from the ovary do not pass through the gills but are washed away from the cloaca by the respiratory current. If there were a reversal of the ciliary beating during ovulation a continuous oozing of eggs through the gills would be expected. Addi- tional negative evidence is supplied by the fact that in several instances spawning oysters have been observed to discharge eggs both through the gills and through the cloaca.
We have already seen that the emergence of eggs through the ostia coincides with the period of maximal gaping of the valves. These facts led the author to consider whether rhythmical shell movement might cause a suction which would draw the eggs from the suprabranchial chamber into the water tubes and pallial cavity (see reference to this
476
PAUL S. GALTSOFF
in Hopkins' paper, 1937, p. 457). Observations on oysters in \vhich the gills were exposed by cutting a piece of shell, thus forcing the mantle to roll up and leave the pallial chamber wide open, make this explanation untenable.
Observations made during the summer of 1937 with sex-inverted oysters (Galtsoff, I1 '37) are of particular interest because they show that something mure than rhythmical shell movement is necessary for the passage of eggs through the gills. In these experiments one adult female, which during the preceding summer was a male, discharged the eggs through the cloaca in male fashion in spite of the well-developed contractions of the adductor which are typical for a spawning female.
LATENT PERIOD
OVULATION
L I L
I .1 L
J L
I I
J L
FIG. 9. Ostrca gigas, spawning reaction of two females. A, beginning of the reaction; note change in t<mn> level during the latent period and rhythmical contractions during ovulation. H. spawning reaction about one hour after its beginning; during the brief period^ of contractions eggs are discharged through the gills; during the relaxation periods (horizontal parts of the curve) they are discharged through the cloaca. Time interval, 1 minute (bottom line).
< )bviously some other conditions necessary to drive the eggs through the i;ills were still wanting.
The fact that the eggs are forced through the gills from the inside indicates that pressure in the suprabranchial chamber is higher than on the other side of the gill lamellae. The increase in pressure can easily be produced by the discharge of a large amount of material from the gonad and closure of the cloacal opening. Although direct evidence is lacking, tbi- supposition is in accord with other observed facts. Since it is known that the discharge of eggs from the ovary once started con- tinues without interruption for a long time, it is logical to expect that pressure inside the suprabranchial chamber increases when the contrac-
SPAWNING REACTIONS OF OSTREA VIRGINICA 477
tion of the adductor closes the valves and compresses the gill lamella?. During the following relaxation and gaping of the valves the gill lamellae spread apart and eggs which had accumulated inside are forced through the water tubes and ostia. This may be possible only if the cloacal opening remains closed for otherwise they would be carried away with the outgoing current of water. Whether the soft and contractile edges of the cloaca come together, closing the opening during ovulation, has not been ascertained by observation but the spreading and contraction of the gill lamellae has often been noticed.
The rather complex and circuitous method of discharging eggs de- scribed for 0. virginica occurs also in other bivalves. It is known that in many incubatory forms the ova are retained in the interlamellar branchial spaces. In Ostrea cdulis, 0. angasi, O. lurida, and Entovalva sp. (Pelseneer, 1906, p. 244) eggs develop in the pallial chamber outside the gills, but in the Unionida they are retained in the interlamellar branchial spaces in which special pouches or marsupia are developed (Ortmann, 1911; Lefevre and Curtis, 1910). Ovulation in the latter forms was observed in but a few cases (Latter, 1891; Ortmann, 1911, p. 298; Howard, 1914, p. 35). Latter, describing the process in Anodonata and Unio, states that ' In order that the ova may reach, their final resting place there must be some reversal of the respiratory currents." He could not, however, detect any reversal of the ciliary motion and attributed the passage of eggs from the cloaca into the gills to a suction caused by the gaping of the valves. This conclusion, how- ever, was not verified by observation.
Stafford (1915) states that in 0. Inrida eggs liberated from gono- ducts into the suprabranchial chambers, being too heavy to be carried in the respiratory current, flow naturally into the water tubes and are forced through them into the pallial chamber by the pressure of their mass. This explanation, apparently not based on observation, requires further verification for it is very doubtful that the ova of this species are much heavier than those of O. virginica. In the latter species, as has been shown above, they are easily carried away by the current. Furthermore, it seems improbable that the weight of eggs will be suf- ficient to counteract the current maintained in the water tubes by the ciliated epithelium.
Spawning in all three groups of bivalves mentioned above may be considered as representing three consecutive stages of one process. In Unionldce, the ova penetrate into the water tubes which develop into special broodchambers, to remain there until the emergence of the glochidia; in the incubatory species of oysters, O. cdulis and O. lurida, the ova completely pass through the gill lamellae but are retained in the
478
PAUL S. GALTSOFF
pallial chamber between the demi-branchs ; in O. virginica, O. gigas and O. commcrcialis, they are discharged from the pallial chamber im- mediately after being forced through the gills. Here the process has reached its final development. All three different types of spawning may be attributed to a greater or lesser efficiency of the combined activity of the adductor muscle, gills, and cloaca during ovulation. A detailed physiological study «f the spawning reactions of various species of Unionidcc and of the incubatory oysters will therefore be interesting.
SEXUAL REACTIONS OF 0. gigas AND 0. commercialis
Sexual reactions in 0. gigas and 0. commercials are similar to those described fur O. I'irginica. Kymograph tracings of shell movements of spawning females of these oysters show the same typical character- istics, namely, change in tonus level during the latent period, rhyth-
LATENT PERIOD
~i — I — i — i — r-
OVULATION
i — i — r
i — r
IMC. 10. Ostrca commercialis, spawning reaction of a female. A, muscular behavior during feeding. B, spawning reaction. Time interval, 1 minute (bot- tom line).
micity in the contractions of the adductor muscle and gradual fading of the reaction (Figs. 9, 10). The reaction in 0. gigas differs slightly, however, from that of the other two species. As a rule, the Japanese oyster discharges eggs both through the gills and cloaca. In one of the experiments, for instance, muscular contractions and discharge through the gills continued for 13 minutes while ovulation and emission of eggs through the cloaca lasted 21 minutes longer. It is impossible to state which of the two methods of discharge is the principal one in this specie- l-'v.gs discharged with the respiratory current through the cloaca often conn- in large lumps, several millimeters long, which im- mediately settle on the bottom. Large numbers of eggs discharged in this manner are smothered and perish while those discharged through the gills are more widely distributed throughout the water and have a better chance to survive. No discharge of eggs through the cloaca was observed in O. commercialis.
SPAWNING REACTIONS OF OSTREA VIRGINICA 479
DURATION AND NUMBER OF SPAWNINGS
Spawning once started may continue for a long time and cannot be stopped by taking the oyster out of the water or even by inflicting
U
D
vl^^^^^A-^^^^^
^
> i i t i i i i i i i i i i i
FIG. 11. Ostrea virginica, muscular behavior of five males (A-E) during ejaculation. Time interval, 1 minute (bottom line).
severe injury to its body. Duration of ovulation, i.e., of the discharge of ova from the gonad, cannot be measured with accuracy since the release of small numbers of eggs can easily be overlooked. Duration of the typical spawning reaction of the adductor, on the other hand,
480 PAUL S. GALTSOFF
can be- determined from the kymograph records, except in the cases where the fading of the curves is so gradual that the cessation of the reaction is indistinct. In many instances, however, the curves of mus- cular activity show changes sufficiently sharp to indicate the end of spawning. In 101 records of 0. virginica available for analysis, the duration of spawning varied from 4 to 118 minutes. The shortest periods were recorded either at the beginning or toward the end of the reproductive season. In 0. gigas (13 records), the duration of the discharge of eggs either through the gills or through the cloaca varied from 15 to 130 minutes. In the two records obtained with 0. coni- incrcialis spawning lasted 12 and 29 minutes.
Copious nvulation was observed always to be accompanied by a large amplitude tif the up and down excursions of the adductor and the greater frc<|uency of its contractions. The greatest number of eggs discharged during one spawning was found to be 114.8 millions in 0. virginica and 55. S millions in O. gigas (Galtsoff, 1930a). In both cases histological examination of the spawned females revealed that only a small portion of the sex cells had been released. Judging by the thickness of the gonad layer, the fertility of 0. gigas appears to be much great tr than that of O. virginica. There is no doubt that the number of eggs discharged during one spawning, given above, does not repre- sent the maximal figure.
That 0. virginica may spawn several times during one season was ascertained by experiments conducted during several summers between June 15 and August 7, at \Yoods Hole. Under laboratory conditions a single female was induced to spawn seven times during this period. How often an individual oyster spawns in nature has not been deter- mined but it may be assumed that under favorable conditions it behaves in its natural environment in the same manner as in the laboratory. Re- peated spawning was .also observed in 0. gigas. In the laboratory the females of this species were induced to spawn two or three times during one mi .nth.
EJACULATION
During ejaculation sperm discharged from the testis passes through the spermiduct and urinogenital cleft into the suprabranchial chamber from whence it is carried by the current of water into the cloaca and washed away. Muscular contractions of the adductor play no role in the release of s],rnn which is effected primarily by the ciliary motion of the epithelial lining of the gonad and respiratory current produced by the gills. The behavior of the adductor during ejaculation shows a great variety of conditions from almost a complete quiescence (Fig.
SPAWNING REACTIONS OF OSTREA VIRGINICA
481
11, ./ ) to greatly increased activity characterized by small and irregular contractions (Fig. 11, C and 11} . These contractions are not, however, concerned with the release of sperm.
The sudden appearance of a jet of sperm at the beginning of ejacula- tion gives an impression of increased velocity of the cloacal current. The question whether there is an increase in ciliary activity during ejaculation was carefully studied in a number of experiments. Two methods were employed for measuring the strength of the cloacal cur-
5.0
STIMULUS
EJACULA TION
4.0
Q
Z
o
o: 3.0
LJ
a.
Q. O oc
0 2.0
1.0
100
200
300
400 500
SECONDS
600
700
800
900
FIG. 12. Rate of pumping of water of the male O. virginica during con- tinuous ejaculation. Electric drop-counting method. Each point represents 10- second average of the number of drops of water discharged by the gills.
rent. In the first series of experiments an electric drop-counting tech- nique, fully described in a previous publication (Galtsoff et al, 1935, p. 172) was used. The method consists in recording on a kymograph each drop of water pumped by the gills through a rubber tubing inserted into the cloaca. The method completely eliminates the effects of the adductor muscle and of the mantle which may constrict the pallial aper- ture, making it possible to obtain measurements of the efficiency of the ciliary epithelium alone. The results of the two experiments in which continuous records were obtained before, during and after ejaculation
482
PAUL S. GALTSOFF
are presented in Figs. 12 and 13. The experiments lasted 14 and 32 minutes respectively. Each point plotted on the curves represents an average number of drops per second for the preceding 10-second inter- val. The exact time of stimulation and of the release of sperm is indi- cated by the arrows. In the first experiment (Fig. 12) ejaculation lasted continuously for several minutes. There was no increase in the ciliary activity at the beginning of ejaculation and, with the exception of one high peak and four depressions, the rate of pumping remained more or less constant. Stimulation, which in this case consisted in adding
O 2-0
z
O
u
LJ ID
:
.CS f
\ \
EJACULATION
EGGS
NO REACTION
100
200
300 400
ADRENALIN
500 600
SEAWATER
700
800
900
1000
EGGS
2-0
n. 1.0
O a
EJACU L AT ION
1
1100
1200
1300
1400 1500
SECONDS
1600
1700
1800
1900
?000
FIG. 13. Rate of pumping of water by the male O. rirginica during and be- tween ejaculations. Electric drop-counting method. Each point represents 10- second average of number of drops of water discharged by the gills.
thyroidin to the pallial cavity of the oyster, was followed by a temporary inhibition of the rate of pumping. This was probably due to me- chanical stimulation, as a similar effect was noticeable in other cases when sea water of the same temperature was injected between the valves (Fig. 13). In the second experiment ejaculation was induced three times by the addition of oyster eggs. Not only was there no increase in the rate of pumping of water, but in two cases (Fig. 13, upper line), the rate materially decreased during ejaculation.
Similar results were obtained by using the technique of Hopkins (1933) for recording the relative strength of the cloacal current of an
SPAWNING REACTIONS OF OSTREA VIRGINICA
intact oyster. In this method the velocity of the current is measured by the deflection of a light lever, one arm of which supports a small paper or celluloid cone placed in front of the cloaca! opening. The deflections of the lever are proportional to the velocity of the current striking the cone. The method permits the study of the operations of the entire complex mechanism which controls the pumping of water by the gills. A photographic reproduction of a portion of one of the records is given in Fig. 14. The exact time of stimulation and the beginning and end of ejaculations are indicated by vertical marks (third line). The upper line represents the shell movement of the spawning male and the changes in the strength of current are shown on the second line. The uppermost points of both curves correspond to the closure
MUSCULAR CONTRACTION
CLOACAL CURRENT
STIMULUS EJACULATION EJACULATION
n h | H
FIG. 14. Muscular contraction and relative strength of cloacal current during ejaculation of O. virginica. Double vertical bars (third line from top) mark the beginning of ejaculation. Single bars indicate its cessation. Time interval, 1 second (bottom line).
of shells and cessation of current. The record clearly shows a correla- tion between the strength of the current and shell movement and lack of significant change in the activity of the ciliated epithelium during ejaculation. The experiments repeated several times under various conditions always gave the same results indicating that sudden discharge of sperm during ejaculation is not due to the outburst of increased ciliary activity. Although no distinct sphincter is present in the spermi- ducts or in the urinogenital cleft, the ability of the male repeatedly to release small amounts of sperm under the influence of a proper stimulus suggests a sphincter-like action. Histological examination provides certain basis for this assumption for it shows (Fig. 3, M) the existence in the walls of the spcrmiducts of muscle fibres which are stronger and more abundant than in the oviducts. It is therefore probable that while the ciliated lining of the genital canals and spermiducts provides a motive
484 PAUL S. GALTSOFF
power which conveys the ripe sperm from the testis, the release of sperm is regulated by the contractions of muscle fibres scattered in the walls of the ducts.
Ejaculation may proceed either from one or simultaneously from both spermiclucts. At the beginning of the spawning season only a small amount of sperm is discharged at each ejaculation, which lasts only a few seconds, but repeated emissions can be induced immediately one after another. In several experiments a single male was stimulated to shed sperm more than a hundred times during a ten-hour period, each time- releasing a small amount of spermatozoa. As the season proceeds and the amount of ripe sperm in the testis increases, each ejaculation lasts longer. Finally a stage is reached when ejaculation, once started, continues for hours and cannot be stopped until the male is completely spent.
Sperm discharged with the cloacal current remains in suspension for a considerable time and is widely distributed in the water.
The males of O. gigas and O. cominercialis behave in the same man- ner as the males of O. virginica.
SUMMARY
The gonad of an oyster consists of a large number of branching tubules containing sex cells and emptying into genital canals, the lumen "I which is lined on one side with the ciliated epithelium. There are two gonoducts, one on each side of the gonad, which open into supra- branchial chambers. Spawning of the female consists in the discharge of eggs from the ovary (ovulation) ; rhythmical contractions of the adductor which cause opening and closing of the shell valves ; and in the contraction and change in the position of mantle borders which almost completely close the pallial cavity. The last two phases of spawning are accessory sexual reactions which provide mechanism for better dispersal of eggs in the water. Muscular behavior during spawn- ing is characterized by its remarkable rhythmicity and maintenance of a constant tonus level. It cannot be reproduced by artificial stimulation, barged into the suprabranchial chamber are not carried away by the cloaca! current but penetrate through the gills into the pallial cavity and are expelled by vigorous movements of the shell valves. rheir ••• through the gills coincides with the periods of greatest
relaxation of the adductor and widest gaping of shell. If the shell movement is prevented by severing the hinge, eggs fail to pass through the gills and are discharged by the cloacal current. The possibility of the reversal of the ciliary motion and of the suction produced by shell
SPAWNING REACTIONS OF OSTREA VIRGINICA 485
movement as the forces responsible for the passage of eggs through the gills are discussed and rejected as untenable. The following tenta- tive explanation is offered. The discharge from the ovary and the closure of the cloaca produce an increased pressure in the suprabranchial chamber which forces the eggs through the water tubes and gill ostia into the pallial cavity. The difference in pressure on both sides of the gill reaches its maximum when the valves open and the gill lamella1 spread apart. This accounts for the penetration of eggs through the gill during the periods of greatest relaxation of the adductor.
Ejaculation is not accompanied by the rhythmical contractions of the adductor and the sperm discharged through the spermiduct is car- ried away by the cloacal current. There is no increase in the rate of pumping of water during spawning.
The discharge of sperm is probably controlled by the contractions of numerous muscle fibres in the walls of the spermiduct. Definite sphincter-like structure is absent.
REFERENCES
COE, W. R., 1936. Environment and sex in the oviparous oyster, Ostrea virginica. Biol. Bull., 71: 353.
GALTSOFF, P. S., 1928. Experimental study of the function of the oyster gills and its bearing on the problems of oyster culture and sanitary control of the oyster industry. Bull. U. S. Bur. of Fisheries, 44: 1.
GALTSOFF, P. S., 1930a. The fecundity of the oyster. Science, 72: 97.
GALTSOFF, P .S., 1930/>. The role of chemical stimulation in the spawning reac- tions of Ostrea virginica and Ostrea gigas. Proc. Nat. Acad. Sci., Washington, 16: 555.
GALTSOFF, P. S., 1937. Observations and experiments on sex change in the adult American oyster, Ostrea virginica (Abstract). Biol. Bull., 73: 356.
GALTSOFF, P. S., H. F. PRYTHERCH, R. O. SMITH, AND V. KOEHRING, 1935. Ef- fects of crude oil pollution on oysters in Louisiana waters. Bull. U. S. Bur. Fisheries, 48 (No. 18) : 143.
HOPKINS, A. E., 1931. Factors influencing the spawning and setting of oysters in Galveston Bay, Tex. Bull. U. S. Bur. of Fisheries, 47: 57.
HOPKINS, A. E., 1933. Experiments on the feeding behavior of the oyster, Ostrea gigas. Jour. Expcr. Zool., 64: 469.
HOPKINS, A. E., 1936. Activity of the adductor muscle in oysters. Phvsiol. Zool., 9: 498.
HOPKINS, A. E., 1937. Experimental observations on spawning, larval develop- ment, and setting in the Olympia oyster, Ostrea lurida. Bull. U. S. Bur. of Fisheries, 48 (No. 23) : 439.
HOWARD, A. D., 1914. Experiments in propagation of fresh-water mussels of the Quadrula group. Re[>t. of the U. S. Commissioner of Fisheries for 1913, Appendix 4, 52 pp.
KELLOGG, J. L., 1892. A contribution to our knowledge of the morphology of Lamellibranchiate mollusks. Bull. U. S. Bur. of Fish Commission for 1890, 10: 389.
LATTER, O. H., 1891. Notes on Anodon and Unio. Proc. of Zool. Soc. London for the year 1891: 52.
486 PAUL S. GALTSOFF
LEFEVRE, G., AND W. C. CURTIS, l('lo. Reproduction and parasitism in the Unionidar. Jour. E.rpcr. Zool., 9: 79.
NELSON, Tn. C, 1936. \VatiT liltration by the oyster and a new hormone effect upon the rate of flow. I' roc. Soc. E.rf>er. Biol. and Mcd., 34: 189.
ORTMANX, A. F., 1911. A monograph of the Najadcs of Pennsylvania. Mem. Carnegie Museum. 4 ( Xo. n) : 279.
PELSENEER, P., 1906. Mollu^ca. Ray Lankester, A treatise on Zoology, Part V: 355 pp.
PRYTHERCII, H. !•".. l''_'S. Investigation of the physical conditions controlling spawning of < md the occurrence, distribution, and setting of oyster
larvae in Mil ford Harbor, Connecticut. Bull, of the Bur. of Fisheries, 44: 429.
ROUGHLEY, T. ('.. l'>33. The life history of the Australian oyster, Ostrea com- mcm.-ilis. Proc. Linncan Soc. of NCZV South Wales. 58: 279.
STAFFORD, J.. I'M 5. The native oyster of British Columbia (Ostrea lurida, Car- penter). Province of British Columbia Report Com. of I-'ish, \car end- in,, Dec. 31, 1914: 100.
YONGE. C. M.. l°2o. Structure and physiology of the organs of feeding and diges- tion in Ostrea edulis. Jour. Marine Biol. Ass'n, N. S., 14: 295.
INDEX
ABRAMOWITZ, A. A., AND R. K. ABRAMOWITZ. On the specificity and related properties of the crus- tacean chromatophorotropic hor- mone, 278.
Achromatic figure, origin and develop- ment, 41.
Ammonia, distribution in waters of Gulf of Maine, 83.
Amoeba proteus (Leidy), contractile vacuole, 306.
JDACTERIA, adsorption of, by marine bottom, 56.
BARNES, T. CUNLIFFE. Experiments on Ligia in Bermuda. V. Further effects of salts and of heavy sea water, 108.
BARTH, L. G. Quantitative studies of the factors governing the rate of regeneration in Tubularia, 155.
Bilaterality, Chaetopterus egg, and en- trance point of spermatozoon, 401.
BRACHET, JEAN. The oxygen consump- tion of artificially activated and fer- tilized Chaetopterus eggs, 93.
CALCIUM ions, effect on volume
changes in Amoeba proteus, 297. CHACE, F. A., JR. See Welsh and Chace,
364. Chaetopterus egg, dorso-ventral pre-
organization of, 395. — , - — , oxygen consumption of, when
artificially activated and fertilized,
93.
-, relation between entrance
point of spermatozoon and bilater- ality of, 401.
Chromatophorotropic hormone in crus- taceans, specificity of, 278.
CLAFF, C. LLOYD. See Kidder and Claff, 178.
Cleavage, effects of ultra-violet radiation
of X 2537A, in sea-urchin eggs, 330.
— , in females and haploid males in
Melittobia chalybii Ashmead, 256.
CLEVELAND, L. R. Longitudinal and
transverse division in two closely related flagellates, 1.
-. Origin and develop-
ment of the achromatic figure, 41.
COE, W. R. Primary sexual phases in the oviparous oyster (Ostrea vir- ginica), 64.
Colpoda cucullus, cytological investiga- tions, 178.
Contractile vacuole in Amoeba proteus (Leidy), 306.
Cyclothone, kidney in, 349.
Cytology of Colpoda cucullus, 178.
TMFFERENTIATION, energetics of,
l^y
99.
Diploids from unfertilized eggs in Habro-
bracon, 247. Division, longitudinal and transverse, in
two closely related flagellates, 1. DOUDOROFF, P. See Sumner and Dou-
doroff, 403. Drosophila pseudoobscura, duration of
life without food, 314.
jgM BRYOLOGY, Chaetopterus egg, entrance point of spermatozoon and bilaterality of, 401.
— , chemical, relation of fat changes to, in sea urchin, 267.
— , dorso-ventral pre-organization of Chaetopterus egg, 395.
Embryonic determination, in Sabellaria
vulgaris, 198, 211. Endomixis, delayed occurrence and total
omission of, in selected lines of
Paramecium aurelia, 76.
— , effect of shorter than normal inter- endomictic intervals on mortality after, 244.
— , mortality after, related to prior interendomictic interval in Para- mecium aurelia, 235.
Eyes of deep-sea crustaceans, Serges- tidae, 364.
changes, relation to general chem- ical embryology of sea urchin, 267.
487
488
INDEX
Flagellates, longitudinal and transverse division in two closely related species, 1.
Food, duration of life without, in Droso- phila pseudoohscura, 314.
FOWLER, COLEEN. See Mast and Fow- ler, 297.
QALTSOFF, PAH S. Physiology of reproduction of Ostrea virginica, I, 461.
Gametogenesis in females and haploid males, Melittobia chalybii Ashmead, 256.
GELBER, JULIUS. The effect of shorter than normal interendomictic inter- vals on mortality after endomixis in Paramecium aurelia, 244.
GIESE, ARTHUR C. The effects of ultra- violet radiation of X 2537A upon cleavage of sea-urchin eggs, 330.
Gulf of Maine, distribution of ammonia in, 83.
LJABROBRACON, diploids from un- fertilized eggs, 247. , -perm inactivation by x-radiation, 253.
HAYES, FREDERICK RONALD. The rela- tion of fat changes to the general chemical embryology of the sea urchin, 267.
Histology of retractor muscles of Thyone briareus Lesueur, 342.
Hormone, chromatophorotropic, in crus- taceans, specificity of, 278.
HOROWITZ, N. H. See Tyler and Horo- witz, 99.
] NACTIVATION of sperm by x-radi- ation in Habrobracon, 253.
Intermediate-winged aphids and time-of- determination theory, 430.
I/ KYS, A\ci-:i. H. See Redfield and
Keys, 83. KIDHI i , ( .1 ORC.E W., AND C. LLOYD
( i \i i . Cytological investigations
of Colpoda cucullus, 178. Kidney, trlm-i, in genus Cyclothone,
349.
J^EAVI'I I . BENJAMIN H. The quan- titali\c M-niral distribution of macrozodplankton in the Atlantic Ocean basin. •>7<>.
Levinseniella cruzi, life history of, 319.
Ligia, effects of salts and heavy sea water, 108.
LILLELAND, OLE. Duration of life with- out food in Drosophila pseudo- obscura, 314.
ACROZOOPLANKTON, quantita- tive vertical distribution in Atlantic Ocean basin, 376.
MAST, S. O. The contractile vacuole in
Amoeba proteus (Leidy), 306. — , - — , AND COLEEN FOWLER.
The effect of sodium, potassium and calcium ions on changes in volume of Amoeba proteus, 297.
MAXWELL, JANE. Inactivation of sperm by x-radiation in Habrobracon, 253.
Melittobia chalybii Ashmead, sex ratio in, 256.
MORGAN, T. H. A reconsideration of the evidence concerning a dorso- ventral pre-organization of the egg of Chaetopterus, 395.
MORGAN, T. H., AND ALBERT TYLER. The relation between entrance point of the spermatozoon and bilaterality of the egg of Chaetopterus, 401.
XJOYIKOFF, ALEX H. Embryonic determination in the annelid, Sabel- laria vulgaris, I, II, 198, 21 1.
QLSON, MAGNUS. The histology of the retractor muscles of Thyone briareus Lesueur, 342.
Ostrea virginica, physiology of repro- duction, 461.
— , primary sexual phases in, 64.
OWEN, BRADFORD B. A simple teleost kidney in t lie genus Cyclothone, 349.
Oxygen consumption, in artificially ac- tivated and fertilized Chaetopterus eggs, 93.
Oyster, physiology of reproduction in,
461.
— , primary sexual phases in, ovipar- ous, 64.
DARAMECIUM aurelia, delay in and omission of endomixis in, 76.
— , effect of interendomictic in- tervals on mortality after endomixis, 244.
-, mortality after endomixis, re-
lated to prior inlrrcmlomirl ic inter- val in, 235.
INDEX
489
PIERSON, BERNICE F. The relation of mortality after endomixis to the prior interendomictic interval in Paramecium aurelia, 235.
Potassium ions, effect on volume changes in Amoeba, 297.
t> EDFIELI), ALFRED C., AND ANCEL B. KEYS. The distribution of am- monia in the waters of the Gulf of Maine, 83.
Regeneration, factors governing rate of, in Tubularia, 155.
Reproduction, physiology of, in Ostrea virginica, 461.
Respiratory metabolism and temperature acclimatization in fishes, 403.
Retractor muscles, histology of, in Thyone briareus Lesueur, 342.
C ABELLARIA vulgaris, embryonic de- termination, 198, 211.
SCHMIEDER, RUDOLF G. The sex ratio in Melittobia chalybii • Ashmead, gametogenesis and cleavage in fe- males and in haploid males (Hy- menoptera: Chalcidoidea), 256.
Sergestidae, eyes of, 364.
Sex, phases, in oviparous oyster, 64. — , ratio, in Melittobia chalybii Ash- mead, 256.
Sodium ions, effect on volume changes in Amoeba, 297.
SONNEBORN, T. M. The delayed occur- rence and total omission of endo- mixis in selected lines of Paramecium aurelia, 76.
SPEICHER, KATHRYN G., AND B. R. SPEICHER. Diploids from unfer- tilized eggs in Habrobracon, 247.
STILES, KARL A. Intermediate-winged aphids and the time-of-determina- tion theory, 430.
SUMMERS, F. M. Some aspects of nor- mal development in the colonial ciliate Zoothamnium alternans, 117. — , - — . Form regulation in
Zoothamnium alternans, 130.
SUMNER, F. B., AND P. DOUDOROFF
Some experiments upon temperature acclimatization and respiratory met- abolism in fishes, 403.
'pEMPERATURE acclimatization and
respiratory metabolism in fishes,
403. Thyone briareus Lesueur, histology of
retractor muscles, 342. Time-of-determination theory and inter-
mediate-winged aphids, 430. Trematode, life history of, from shore
birds Limosa fedoa and Catop-
trophorus semi-palmatus inornatus,
319. Tubularia, regeneration rate, factors
governing, 155. TYLER, ALBERT. See Morgan and Tyler,
401. — , - - AND N. H. HOROWITZ. On
the energetics of differentiation, VII,
99.
yARTIOVAARA, UNTO. See Waks-
man and Vartiovaara, 56. Volume changes in Amoeba proteus, as
affected by various salts, 297.
, SELMAN A., AND UNTO VARTIOVAARA. The adsorption of bacteria by marine bottom, 56. WELSH, J. H., AND F. A. CHACE, JR. Eyes of deep-sea crustaceans, II. Sergestidae, 364.
V-RADIATION, and inactivation of sperm in Habrobracon, 253.
yOUNG, R. T. The life history of a
trematode (Levinseniella cruzi?)
from the shore birds (Limosa fedoa
and Catoptrophorus semipalmatus
inornatus), 319.
yOOTHAMNILIM alternans, normal development and form regulation, 117, 130.
Biological Materials
The Supply Department of the Marine Rio- logical Laboratory has a complete stock of ex- cellent material for summer school work in Zoology, Botany, Embryology, and Histology.
At the present time a group of skilled prepar- ators is making a fresh collection of marine specimens for the coming year. \Ye shall be very glad to quote prices on any of this mate- rial, which is listed in our current catalogue.
All our materials are guaranteed to give ab- solute satisfaction.
Catalogues are furnished on request
Supply Department
MARINE BIOLOGICAL LABORATORY
Woods Hole, Mass.
Volume LXXIV
Number 1
THE
BIOLOGICAL BULLETIN
PUBLISHED BY
THE MARINE BIOLOGICAL LABORATORY
Editorial Board
GARY N. CALKINS, Columbia University E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University
E. E. JUST, Howard University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden T. H. MORGAN, California Institute of Technology G. H. PARKER, Harvard University EDMUND B. WILSON, Columbia University
ALFRED C. REDFIELD, Harvard University Managing Editor
FEBRUARY, 1938
Printed and Issued by
LANCASTER PRESS, Inc.
PRINCE & LEMON STS. LANCASTER, PA.
BLAKISTON BOOKS for BIOLOGISTS Lee — The Microtomist's Vade Mecum
10th Edition Revised
This new edition of Bolles Lee's laboratory handbook contains many additions and improvements. One important change has been the in- clusion of plant technique, an entirely new section written by Dr. D. G. Catcheside of London University. The book is essential to the biologist, microscopist, botanist and pathologist as it presents the best of the methods in general use and gives definite instructions for every detail of procedure. Prepared by Collaboration of Several Authors of Special Experience in each subject. Edited by J. BRONTE GATENBY, M.A., PH.D., D.Sc. < Trinity College, Dublin), AND THEOPHILUS S. PAINTER, A.M., PH.D. (Texas University).
11 Illustrations. 784 Pages. Washable Cloth $9.00.
Barnes -^Textbook of General Physiology
This new text is comprehensive yet concise and treats the subject in an interesting and adequate manner for college students. Its simplified literary style has been described as " telegraphic — clear — excellent in tone —and of great service to beginning students." It is based upon a year's course in the general physiology of cold-blooded animals. 166 Illustrations. Bibliography. Appendix. 554 Pages. Washable Covers $4.50.
By T. CUNLIFFE BARNES, D.Sc., Assistant Professor of Biology, Yale University.
Imms — Recent Advances in Entomology, 2d Ed.
1 hi1- edition records considerable new progress. By A. D. IMMS, D.Sc., F.R.S. (Cambridge Univ., Eng.). 94 Illus. 431 Pages. Cloth $5.00.
Darlington — Recent Advances in Cytology, 2d Ed.
Contains 160 Illus. 16 Plates. Numerous Tables, Appendices, Bibliog- graphy. 671 Pages. Cloth $6.00. By C. D. DARLINGTON, D.Sc., PH.D. (Loud.).
Glasstone — Recent Advances in Physical Chemistry, 3d Ed.
Presents highly important new developments. 31 Illus. 477 Pages. ' loth S5.00. By SAMUEL GI.ASSTOXK. 1'n.D., Sc.D. (Loud.), F.I.C.
Cameron — Recent Advances in Endocrinology, 3d Ed.
Xcw work has been done in all branches of the subject. 65 Illus. 3 1 'lutes. 458 Pages. Cloth $5.00. By A. T. CAMERON, F.I.C., F.R.S.C. 1 li'innipcg).
Evans — Recent Advances in Physiology, 5th Ed.
An interesting discussion of recent progress. 120 Illus. 500 Pages. $5.00. By C. Lovatt Evans, Revised by W. H. NEWTON, M.D. (Lond.).
P. BLAKISTON'S SON & CO, INC., PHILADELPHIA, PA.
DISSECTING SETS
This illustrates one of the many dissecting sets which comprise cur complete stock. Our catalog number 38 describes and il- lustrates ten sets varying from a simple set for the student to an elaborate one for the research technician. We will gladly send you a catalog upon request.
Also the Largest Va riety of
DISSECTING INSTRUMENTS — MICRO SLIDES and COVER GLASSES - MICRO SLIDE BOXES and TRAYS MAG- NIFIERS- INSECT PINS-- RIKER MOUNTS - MUSEUM JARS PETRI DISHES RUBBER TUBING, ETC.
There are also available separate catalogs on Charts, Models, Specimens and Preparations covering the fields of: Human and Comparative Anatomy, Physiology, Neurology, Zoology, Botany, Embryology, Entomology, Ecology, etc.
THE BIOLOGICAL BULLETIN
THE BIOLOGICAL BULLETIN is issued six times a year. Single numbers, $1.75. Subscription per volume (3 numbers), $4.50.
Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain : Wheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W.C. 2.
Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Mass., between June 1 and October 1 and to the Institute of Biology, Divinity Avenue, Cambridge, Mass., during the remainder of the year.
INSTRUCTIONS TO AUTHORS
Preparation of Manuscript. In addition to the text matter, manuscripts should include a running page head of not more than thirty-five letters. Footnotes, tables, and legends for figures should be typed on separate sheets.
Preparation of Figures. The dimensions of the printed page (4%x7 inches) should be borne in mind in preparing figures for publication. Draw- ings and photographs, as well as any lettering upon them, should be large enough to remain clear and legible upon reduction to page size. Illustrations should be planned for sufficient reduction to permit legends to be set below them. In so far as possible, explanatory matter should be included in the legends, not lettered on the figures. Statements of magnification should take into account the amount of reduction necessary. Figures will be reproduced as line cuts or halftones. Figures intended for reproduction as line cuts should be drawn in India ink on white paper or blue-lined coordinate paper. Blue ink will not show in reproduction, so that all guide lines, letters, etc. must be in India ink. Figures intended for reproduction as halftone plates should be grouped with as little waste space as possible. Drawings and lettering for halftone plates should be made directly on heavy Bristol board, not pasted on, as the outlines of pasted letters or drawings appear in the reproduction unless removed by an expensive process. Methods of repro- duction not regularly employed by the Biological Bulletin will be used only at the author's expense. The originals of illustrations will not be returned except by special request.
Directions for Mailing. Manuscripts and illustrations should be packed flat between stiff cardboards. Large charts and graphs may be rolled and sent in a mailing tube.
Reprints. Authors will be furnished, free of charge, one hundred re- prints without covers. Additional copies may be obtained at cost.
Proof. Page proof will be furnished only upon special request. When cross-references are made in the text, the material referred to should be marked clearly on the galley proof in order that the proper page numbers may be supplied.
Entered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894.
LIVE MARINE AQUARIA SETS
WITH the successful practise for the past seven years of shipping balanced Marine Aquaria Sets, the Supply Department has made it possible for the inland student to have the advantages of those working near the coast where living material is readily available.
The service of supplying live specimens to schools located at some distance from the sea has proven very satisfactory, and successful shipments as far west as the Mississippi and as far south as Georgia are guaranteed from No- vember through February. Our records show that these live aquaria sets have arrived in good condition in New Mexico, Texas, Missouri and other western states, but shipments to such dis- tant points are not guaranteed.
These Marine Aquaria Sets are listed in our new 1937 catalogue in which prices on both living and preserved specimens have in many cases been greatly reduced.
Supply Department
MARINE BIOLOGICAL LABORATORY
Est. 1890
Circular will be gladly sent on request.
The Standard for Microscope Gloss
Gold Seal Microscope Slides and Cover Glasses
CrystalClear • Non-Corrosive • WillNotFog
Gold Seal Slides and Cover Glasses attain a precise uniformity of thin- ness and plane surface that is unpar- alleled. They are brilliantly crystal clear and guaranteed against corro- sion, fogging or any imperfection.
CLAY- ADAMS CO.,1"?
25 EAST 26™ STREET, NEW YORK
A Perfect Illustration
Or the lack of it, may make or mar a scientific paper.
For 65 years we have specialized in making reproductions by the Helio- type process of the most delicate pencil and wash drawings and photo- graphs; and by the Heliochromc proc- ess, of paintings and drawings in color.
Examples of our work may be seen on pp. 26-40 of this issue.
Ask the editor to whom you submit your next paper to secure our esti- mates for the reproduction of your illustrations.
The Heliotype Corporation
Eat. 1872
172 Green St., Jamaica Plain, Boston, Mass.
LANCASTER PRESS, Inc.
LANCASTER, PA.
THE EXPERIENCE we have gained from printing some sixty educational publica- tions has fitted us to meet the standards of customers who demand the best.
We shall be happy to have workers at
the MARINE BIOLOGICAL LABORATORY
write for estimates on journals or monographs. Our prices are moderate.
erimen
tal
biology
^Monographs
unique cumulative reference library for biologists, providing full accounts of both the experimental and the theoretical aspects of all significant current research work in the various biological sciences.
PUBLISHED AND FORTHCOMING VOLUMES
AUTONOMIC NEURO-EFFECTOR SYSTEMS
BY WALTER B. CANNON AND ARTURO ROSENBLUETH. $4.00
PHYTOHORMONES
BY F. W. WENT AND KENNETH V. THIMANN. $4.00
PACEMAKERS IN RELATION TO ASPECTS OF BEHAVIOR
BY HUDSON HOAGLAND. $3.00
NEUROEMBRYOLOGY, AN EXPERIMENTAL STUDY
BY SAMUEL R. DETWILER. $3.75
THE EGGS OF MAMMALS
BY GREGORY PINCUS. $3.75
THE RECEPTOR PROCESS IN VISION
BY SELIG HECHT
GEOTROPISM: A STUDY OF DETERMINISM IN BEHAVIOR
BY W. J. CROZIER AND GREGORY PINCUS
THE BIOLOGY AND CHEMISTRY OF OVARIAN HORMONES
BY GEORGE W. CORNER AND WILLARD M. ALLEN
BIOLOGICAL OXIDATIONS
BY E. S. GUZMAN BARRON
THE HYPOPHYSIS
BY J. B. COLLIP, D. L. THOMSON, AND H. SELYE
NITROGEN METABOLISM IN ANIMALS
BY HENRY BORSOOK
TEMPERATURE CHARACTERISTICS
BY W. J. CROZIER
CYTO-GENETICS AND PLANT PHYLOGENY
BY ERNEST B. BABCOCK
BIOELECTRIC PHENOMENA IN PLANTS
BY L. R. BLINKS
PERIODICITY IN ANIMAL BEHAVIOR
BY T. J. B. STIER
THE GENETIC ANALYSIS OF SPECIES DIFFERENCES
BY T. H. DOBZHANSKY
DYNAMICS OF INFLAMMATION: AN INQUIRY INTO THE MECH- ANISM OF INFECTIOUS PROCESSES
BY VALY MEN KIN
The Macmillan Company
60 Fif ih Avenue
New York
CONTENTS
Page CLEVELAND, L. R.
Longitudinal and Transverse Division in Two Closely Related Flagellates 1
CLEVELAND, L. R.
Origin and Development of the Achromatic Figure 41
WAKSMAN, SELMAN A. AND UNTO VARTIOVAARA
The Adsorption of Bacteria by Marine Bottom 56
COE, W. R.
Primary Sexual Phases in the Oviparous Oyster (Ostrea virgmica) 64
SONNEBORN, T. M.
The Delayed Occurrence and Total Omission of Endomixis
in Selected Lines of Paramecium aurelia 76
REDFIELD, ALFRED C. AND ANCEL B. KEYS
The Distribution of Ammonia in the Waters of the Gulf of Maine 83
BRACKET, JEAN
The Oxygen Consumption of Artificially Activated and Fertilized Chaetopterus Eggs 93
TYLER, ALBERT AND N. H. HOROWITZ
On the Energetics of Differentiation. VII .... 99
BARNES, T. CUNLIFFE
Experiments on Ligia in Bermuda. V. Further effects of salts and of heavy sea water 108
SUMMERS, F. M.
Some Aspects of Normal Development in the Colonial Ciliate Zoothamnium alternans 117
SUMMERS, F. M.
Form Regulation in Zoothamnium alternans. . . 130
Volume LXXIV
Number 2
THE
BIOLOGICAL BULLETIN
PUBLISHED BY
THE MARINE BIOLOGICAL LABORATORY
Editorial Board
GARY N. CALKINS, Columbia University E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University
E. E. JUST, Howard University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden T. H. MORGAN, California Institute of Technology G. H. PARKER, Harvard University EDMUND B. WILSON, Columbia University
ALFRED C. REDFIELD, Harvard University Managing Editor
APRIL, 1938
Printed and Issued by
LANCASTER PRESS, Inc.
PRINCE & LEMON STS.
LANCASTER, PA.
BLAKISTON BOOKS
Textbook of General Physiology
by
T. CUXLIFFK BARXKS. D.Sc.
Professor of Kioloyy, Yale I'n'rccrsit v
A distinctive review from The Journal of The American Medical Association, issue of December 25, 1937.
" Barnes has given a new and fresh point of view in general physiology. Since the publication of Bayliss's Principles of General Physiology, this is the first book in English to give a significantly improved ap- proach. The author himself is a prolific contributor to the field of general physiology, especially in the fundamental problems of water metabolism, and his chapters on physical chemistry applied to physiol- ogy are especially valuable. Of 477 pages, 267 are devoted to such topics. There has not been a more intelligent or more comprehensive biologic treatment of these questions in any language. There is a large and valuable bibliography, and the book is well indexed. The illus- trations are ingenious and many, particularly the diagrammatic ones, are original."
Select this book for your course in General Physiology. Many teachers are finding it stimulating and teachable.
166 Illustrations 477 Text Pages (1937) List Price $4.50
Order ----------------------------------------
P. BlakJStOd'S Son & Co.T Inc., 1012 Walnut Street, Philadelphia
Si-ml mi- :, <«>|.v of Harm's1 (iKNKItA I. I'll \ SIO !.()<; V (Sl.:><»
Vimt' \ililn-ss Si I I
I'.iol. r.nl. I 38
No. A-196
DISSECTING SETS
This illustrates one of the many dissecting sets which comprise our complete stock. Our NEW catalog No. 125 describes and illustrates twelve sets varying from a simple set for the elemen- tary student to an elaborate one for the research technician. We will gladly send you a catalog upon request.
Also the Largest Variety of
DISSECTING INSTRUMENTS— MICRO SLIDES and COVER GLASSES MICRO SLIDE BOXES and TRAYS MAG- NIFIERS - - INSECT PINS - - RIKER MOUNTS - MUSEUM JARS PETRI DISHES RUBBER TUBING, ETC.
There are also available separate catalogs on Charts, Models, Specimens and Preparations covering the fields of: Human and Comparative Anatomy, Physiology, Neurology, Zoology, Botany, Embryology, Entomology, Ecology, etc.
25 EAST 26™ STREET, NEW YORK
• RARY
THE BIOLOGICAL BULLETIN
THE BIOLOGICAL BULLETIN is issued six times a year. Single numbers, $1.75. Subscription per volume (3 numbers), $4.50.
Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain : Wheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W.C. 2.
Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory, Woods Hole, Mass, between June 1 and October 1 and to the Biological Labora- tories, Divinity Avenue, Cambridge, Mass., during the remainder of the year.
INSTRUCTIONS TO AUTHORS
Preparation of Manuscript. In addition to the text matter, manuscripts should include a running page head of not more than thirty-five letters. Footnotes, tables, and legends for figures should be typed on separate sheets.
Preparation of Figures. The dimensions of the printed page (4%x7 inches) should be borne in mind in preparing figures for publication. Draw- ings and photographs, as well as any lettering upon them, should be large enough to remain clear and legible upon reduction to page size. Illustrations should be planned for sufficient reduction to permit legends to be set below them. In so far as possible, explanatory matter should be included in the legends, not lettered on the figures. Statements of magnification should take into account the amount of reduction necessary. Figures will be reproduced as line cuts or halftones. Figures intended for reproduction as line cuts should be drawn in India ink on white paper or blue-lined coordinate paper. Blue ink will not show in reproduction, so that all guide lines, letters, etc. must be in India ink. Figures intended for reproduction as halftone plates should be grouped with as little waste space as possible. Drawings and lettering for halftone plates should be made directly on heavy Bristol board, not pasted on, as the outlines of pasted letters or drawings appear in the reproduction unless removed by an expensive process. Methods of repro- duction not regularly employed by the Biological Bulletin will be used only at the author's expense. The originals of illustrations will not be returned except by special request.
Directiotis for Mailing. Manuscripts and illustrations should be packed flat between stiff cardboards. Large charts and graphs may be rolled and sent in a mailing tube.
Reprints. Authors will be furnished, free of charge, one hundred re- prints without covers. Additional copies may be obtained at cost.
Proof. Page proof will be furnished only upon special request. When cross-references are made in tin- text, the material referred to should be marked clearly on the galley proof in order that the proper page numbers may be supplied.
Filtered October 10, 1902, at Lancaster, Pa., as second-class matter under Act of Congress of July 16, 1894.
Biological Materials
For fifty years the Supply Department of the Marine Biological Laboratory has been giving1 dependable service in the furnishing' of living and preserved specimens.
Its staff of collectors and preparators has had many years of experience. These are the fac- tors which enter into the collection and prep- aration of marine specimens and they explain in a large measure the uniformly good quality of our materials.
Catalogues are furnished on request
Supply Department
MARINE BIOLOGICAL LABORATORY
Woods Hole, Mass.
Circular will gladly be sent on request.
A Perfect Illustration
Or the lack of it, may make or mar a scientific paper.
For 65 years we have specialized in making reproductions by the Helio- type process of the most delicate I>encil and wash drawings and photo- graphs; and by the I leliochrome proc- ess, of paintings and drawings in color.
Examples of our work may be seen on pp. 190-192 of this issue,
Ask the editor to whom you submit your next paper to secure our esti- mates for the reproduction of your illustrations.
I he Heliotype Corporation
Eat. 1872
172 Green St., Jamaica Plain, Boston, Mass.
Tfie Standard for Microscope Glass
Gold Seal Microscope Slides and Cover Glasses
Crystal Clear- Non-Corrosive- Will Not Fog
Gold Seal Slides and Cover Glasses attain a precise uni- formity of thinness and plane surface that is unparalleled. They are brilliantly crystal clear and guaranteed against corro- sion, fogging or any imperfec- tion. Microscopic work de- serves the best — specify Gold Seal Slides and Cover Glasses.
25 EAST 26™ STREET. NEW YORK
LANCASTER PRESS, Inc.
LANCASTER, PA.
THE EXPERIENCE we have gained from printing some sixty educational publica- tions has fitted us to meet the standards of customers who demand the best.
We shall be happy to have workers at
the MARINE BIOLOGICAL LABORATORY
write for estimates on journals or monographs. Our prices are moderate.
Forthcoming Biology Texts
Animal Biology
l\c:'iscd lidilion
Bv L. L. WOODRUFF
The new edition of this well known text includes a now rhuptn mi the Human Background, discussing prehistoric man. The whole book has been brought up to date, with careful revision of such topics as genetics on which fruitful research has recently been done. To be published in March. $3.75 (probable).
Genetics
Fourth Edition
By H. E. WALTER
Professor Walter has largely rewritten this introductory text in order to take account of the many new developments in the science of genetics in recent years. There is also a new section in this edition on Human Conservation which discusses some of the important problems of eugenics. To be published this spring. $3.25 (probable).
An Introduction to Vertebrate Anatomy
By H. M. MESSER
The material in this new text, designed for the one-semester course, is presented under anatomical systems. A general discus- sion of each organ system is followed by a comparative study of the system in various type forms. The whole book is clearly and interestingly written, with full material on all essentials but not too much to be covered in the usual half-year course. To be published in May. $3.50 (probable).
Manual for Comparative
By L. P. SAYLES
Organized on the comparative plan, this manual provides direc- tions for study of each anatomical system in all of the commonly used types — amphioxus, squalus, necturus, turtle, pigeon, cat — and in others in less detail. Space is allowed for student drawings and notes so that the book may serve as a workbook as well as a guide. To be published in May.
MACMILLAN - NEW YORK
EARTH, L. G.
CONTENTS
Page
Quantitative Studies of the Factors Governing the Rate of
Regeneration in Tubularia 155
KIDDER, GEORGE W. AND C. LLOYD CLAFF
Cytological Investigations of Colpoda cucullus 178
NOVIKOFF, ALEX B.
Embryonic Determination in the Annelid, Sabellaria vulgaris.
1 198
NOVIKOFF, ALEX B.
Embryonic Determination in the Annelid, Sabellaria vulgaris.
II 211
PIERSON, BERNICE F.
The Relation of Mortality after Endomixis to the Prior Inter-
endomictic Interval in Paramecium aurelia 235
GELBER, JULIUS
The Effect of Shorter than Normal Interendomictic Intervals
on Mortality after Endomixis in Paramecium aurelia 244
SPEICHER, KATHRYN G. AND B. R. SPEICHER
Diploids from Unfertilized Eggs in Habrobracon 247
MAXWELL, JANE
Inactivation of Sperm by X-Radiation in Habrobracon 253
SCHMIEDER, RUDOLF G.
The Sex Ratio in Melittobia chalybii Ashmead, Gameto- genesis and Cleavage in Females and in Haploid Males (Hymenoptera : Chalcidoidea) 256
HAYES, FREDERICK RONALD
The Relation of Fat Changes to the General Chemical Em- bryology of the Sea Urchin 267
ABRAMOWITZ, A. A. AND R. K. ABRAMOWITZ
On the Specificity and Related Properties of the Crustacean Chromatophorotropic Hormone 278
MAST, S. O. AND COLEEN FOWLER
The Effect of Sodium, Potassium and Calcium Ions on Changes in Volume of Amoeba proteus 297
MAST, S. O.
The Contractile Vacuole in Amoeba proteus (Leidy) 306
LlLLELAND, OLE
Duration of Life without Food in Drosophila pseudoobscura. 314
YOUNG, R. T.
The Life History of a Trematode (Levinseniella cruzi?) from the Shore Birds (Limosa fedoa and Catoptrophorus semi- palmatus inornatusj 319
GIESE, ARTHUR C.
The Effects of Ultra-Violet Radiation of X 2537A upon Cleav- age of Sea-Urchin Eggs 330
OLSON, MAGNUS
The Histology of the Retractor Muscles of Thyone briareus Lesueur 342
Volume LXXIV
Number 3
THE
BIOLOGICAL BULLETIN
PUBLISHED BY
THE MARINE BIOLOGICAL LABORATORY
Editorial Board
GARY N. CALKINS, Columbia University E. G. CONKLIN, Princeton University E. N. HARVEY, Princeton University SELIG HECHT, Columbia University LEIGH HOADLEY, Harvard University M. H. JACOBS, University of Pennsylvania H. S. JENNINGS, Johns Hopkins University
E. E. JUST, Howard University FRANK R. LILLIE, University of Chicago CARL R. MOORE, University of Chicago GEORGE T. MOORE, Missouri Botanical Garden T. H. MORGAN, California Institute of Technology G. H. PARKER, Harvard University EDMUND B. WILSON, Columbia University
ALFRED C. REDFIELD, Harvard University Managing Editor
JUNE, 1938
Printed and Issued by
LANCASTER PRESS, Inc.
PRINCE & LEMON STS.
LANCASTER, PA.
BLAKISTON BOOKS
The Microtomist's Vade-Mecum— 10th Ed.
LEE
This standard ami convenient reference book for the laboratory mnv contains the methods of both animal and plant microscopic anatomy. The revision has been very extensive and the book has been considerably enlarged. Special water- anil vermin-proof fabric has been used for the binding.
i«tli F.dition. ii Illtis. 784 Pages. Washable Fabric $9.00
Prepared by (.'ollaboration of Several Scientists; Edited by Dr. J. R. GATKXBY, Trinity College (Dublin) and PROF. T. S. PAINTER (Texas Univ.)
Manual of Common Invertebrate Animals
Revised Edition— PRATT
This is a comprehensive manual (exclusive of insects). Analytical tables and de- scriptions of species are preceded by a morphological description of the group together with an account of habits and distribution of the animals. Teachers and students of the invertebrates will find the book of great reference value.
974 Illus. 854 Pages. Washable Fabric $7.50
By H. S. PRATT, PH.D. (Haverford College)
Comparative Anatomy— NEAL and RAND
' The presentation is clear and the idea is not buried under a blizzard of facts. Students will find it a source of knowledge and understanding. In the light thrown by this book on comparative anatomy, facts glow with significance and interest."
sjn Illus., 4 In Colors. Glossary. 739 Pages. {Cashable Fabric $-f./5 By H. V. XF.AI., PH.D. (Tufts) and H. W. RAND, Pn.D. (Harvard)
Textbook of General Physiology
BARNES
" Barnes has given a new and fresh point of view to general physiology. Since the publication of Bayliss's Principles, this is the first book in English to give a significantly improved approach. There is a large and valuable vocabulary and the book is well indexed. The illustrations are ingenious and many, particularly the diagrammatic ones, are original."
766 Illus. Bibliography. 554 Pages, li'ashablc Fabric $-f.;>o
By T. CUNLIFFF. BARNES, D.Sc. (Yale Univ.)
Introduction and Guide to the Study of Histology— LAMBERT
This new book, an expansion of the author's former successful text, provides a course primarily for students whose objective is not to become histologists but to ;ir<|iiire a knowledge of the subject as a foundation for further study of physiology and pathology or to add a general survey of the subject to a more or less extended course in biology.
flltis. 5/' Pages. Washable Fabric #5.00
P.y AvfHY it. I.AMBKRT, PH.D. (Stale University of
Order ----------------------------------------
P. BlakJSton's Son & Co., Inc., 1012 Walnut Street, Philadelphia
Please send, and rh;irno my :ic< omit, the following hooks:
N iinir Addrt-HH
». H. June 38
No. A-196
DISSECTING SETS
This illustrates one of the many dissecting sets which comprise our complete stock. Our NEW catalog No. 125 describes and illustrates twelve sets varying from a simple set for the elemen- tary student to an elaborate one for the research technician. We will gladly send you a catalog upon request.
Also the Largest Va riety of
DISSECTING INSTRUMENTS— MICRO SLIDES and COVER GLASSES - MICRO SLIDE BOXES and TRAYS MAG- NIFIERS - INSECT PINS - - RIKER MOUNTS - MUSEUM JARS PETRI DISHES RUBBER TUBING, ETC.
There are also available separate catalogs on Charts, Models, Specimens and Preparations covering the fields of: Human and Comparative Anatomy, Physiology, Neurology, Zoology, Botany, Embryology, Entomology, Ecology, etc.
THE BIOLOGICAL BULLETIN
THE BIOLOGICAL BULLETIN is issued six times a year. Single numbers, $1.75. Subscription per volume (3 numbers), $4.50.
Subscriptions and other matter should be addressed to the Biological Bulletin, Prince and Lemon Streets, Lancaster, Pa. Agent for Great Britain: Wheldon & Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W.C. 2.
Communications relative to manuscripts should be sent to the Managing Editor, Marine Biological Laboratory. Woods Hole, Mass., between June 1 and October 1 and to the Biological Lab- oratorio. Divinity Avenue. Cambridge, Mass., during the re- maindrr of the year.
INSTRUCTIONS TO AUTHORS
Preparation of Manuscript. In addition to the text matter, manuscripts should include a running page head of not more than thirty-five letters. Footnotes, tables, and legends for figures should be typed on separate sheets.
Preparation of Figures. The dimensions of the printed page (4% x 7 inches) should be borne in mind in preparing figures for publication. Draw- ings and photographs, as well as any lettering upon them, should be large enough to remain clear and legible upon reduction to page size. Illustrations should be planned for sufficient reduction to permit legends to be set below them. In so far as possible, explanatory matter should be included in the legends, not lettered on the figures. Statements of magnification should take into account the amount of reduction necessary. Figures will be reproduced as line cuts or halftones. Figures intended for reproduction as line cuts should be drawn in India ink on white paper or blue-lined coordinate paper. Blue ink will not show in reproduction, so that all guide lines, letters, etc. must be in India ink. Figures intended for reproduction as halftone plates should be grouped with as little waste space as possible. Drawings and lettering for halftone plates should be made directly on heavy Bristol board, not pasted on, as the outlines of pasted letters or drawings appear in the reproduction unless removed by an expensive process. Methods of repro- duction not regularly employed by the Biological Bulletin will be used only at the author's expense. The originals of illustrations will not be returned except by special request.
Directions for Mailing. Alanuscripts and illustrations should be packed flat between stiff cardboards. Large charts and graphs may be rolled and sent in a mailing tube.
Reprints. Authors will be furnished, free of charge, one hundred re- prints without covers. Additional copies may be obtained at cost.
Proof. Page proof will be furnished only upon special request. When cross-references are made in the text, the material referred to should be marked clearly on the galley proof in order that the proper page numbers may be supplied.
F.ntrn-.' r 10, 1002, at Lancaster, Pa., as second-class matter under
Act of Congress of July 16, 1894.
THE RANGE OF YOUR MICROSCOPICAL WORK
3 » * .4 * I CAN BE EXTENDED WITH
B & L ACCESSORIES
B & L Microscope Accessories will extend the range of your
present equipment to more advanced work at minimum cost. Whether it is new eyepieces, objectives or condensers, or any one of the special accessories needed to adapt the microscope to specialized research or laboratory work you will obtain the greatest added usefulness from your present equipment with B & L Microscope Accessories. Every item is manufactured to the highest standards of accuracy and precision. Write for Catalog D-184. Bausch & Lomb Optical Co., 707 St. Paul Street, Rochester, N. Y.
BAUSCH 6- LOMB
. ... WE MAKE OUR OWN GLASS TO INSURE STANDARDIZED PRODUCTION
FOR YOUR GLASSES INSIST ON B A L ORTHOGON LENSES AND B & L FRAMES . . .
IMPORTANT NOTICE TO SUBSCRIBERS
IBRARIES and individuals desiring to complete sets or uns of the BIOLOGICAL BULLETIN will be able to ob- tain certain numbers and volumes at reduced prices. In order to equalize stock, a number of special offers are being made, including reduced prices on a limited number of copies of the Index to the first sixty volumes of the Bulletin.
Prices will be furnished on request. Please address communications to :
SECRETARY,
THE BIOLOGICAL BULLETIN, Marine Biological Laboratory, Woods Hole, Massachusetts
Circular will gladly be sent on request.
The Standard for Microscope Glass
Gold Seal Microscope Slides and Cover Glasses
Crystal Clear -Non-Corrosive -Will Not Fog
Gold Seal Slides and Cover Glasses attain a precise uni- formity of thinness and plane surface that is unparalleled. They are brilliantly crystal clear and guaranteed against corro- sion, fogging or any imperfec- tion. Microscopic work de- serves the best — specify Gold Seal Slides and Cover Glasses.
CLAY- ADAM SCO,,1
25 EAST 26™ STREET, NEW YORK
A Perfect Illustration
Or the lack of it, may make or mar a scientific paper.
For 65 years we have specialized in making reproductions by the Helio- type process of the most delicate pencil and wash drawings and photo- graphs; and by the Heliochrome proc- ess, of paintings and drawings in color.
Ask the editor to whom you submit your next paper to secure our esti- mates for the reproduction of your illustrations.
1 he Heliotype Corporation
E»t. 1872
172 Green St., Jamaica Plain, Boston, Mass.
LANCASTER PRESS, Inc.
LANCASTER, PA.
THE EXPERIENCE we have gained from printing some sixty educational publica- tions has fitted us to meet the standards of customers who demand the best.
We shall be happy to have workers at
the MARINE BIOLOGICAL LABORATORY
write for estimates on journals or monographs. Our prices are moderate.
IILLA1V'
MA
New Publications in the Field of the Biological Sciences
AN INTRODUCTION TO VERTEBRATE
ANATOMY. By Harold
M. Messer. (June) $3.50, probable.
Presents the essentials for a one-se; mester course, organized under ana- tomical systems rather than under types. The illustrations are outstand- ing.
MANUAL FOR COMPAR- ATIVE ANATOMY
By Leonard P. Sayles. (May) $1.75, probable.
Full directions for the study of all animals commonly used in laboratory work on comparative anatomy, given under each of the organ systems. Am- ple space for student drawings and notes.
A LABORATORY AND FIELD GUIDE TO
BIOLOGY to accompany
THE LIVING WORLD
By Samuel H. Williams. (May) $1.25, probable.
An excellently planned manual for general cultural courses in biology or nature study, prepared especially to ac- company the author's text and empha- sizing the study of living specimens in their natural habitats.
ANIMAL BIOLOGY
Revised Edition.
By L. L. Woodruff. $3.75
In the new edition of this popular text all material has been brought up-to-date and a new chapter added on the Human Background surveying the prehuman Irneage.
GENETICS. 4th Edition.
By Herbert E. Walter. (May 24th) $3.25, probable.
This new edition, rewritten in the light of 25 years of teaching experience and brought up-to-date, presents an out- standingly interesting introduction to genetics.
BACTERIOLOGY
4th Edition.
By E. D. and R. E. Buchanan. (Sept.) $3.50, probable.
A standard text for beginning students, revised and brought up-to-date.
LABORATORY DIREC- TIONS IN COLLEGE
ZOOLOGY. Revised Edition. By H. L. Bruner. $1.75
Revised to coordinate with the new 4th Edition of Hegner's College Zool- ogy.
The Macmillan Company • 60 Fifth Avenue • New York
CONTENTS
Page OWEN, BRADFORD B.
A Simple Teleost Kidney in the Genus Cyclothone 349
WELSH, J. H., AND F. A. CHACE, JR.
Eyes of Deep-Sea Crustaceans. II. Sergestidae 364
LEAVITT, BENJAMIN B.
The Quantitative Vertical Distribution of Macrozooplankton
in the Atlantic Ocean Basin 376
MORGAN, T. H.
A Reconsideration of the Evidence Concerning a Dorso- ventral Pre-organization of the Egg of Chaetopterus 395
MORGAN, T. H., AND ALBERT TYLER
The Relation between Entrance Point of the Spermatozoon and Bilaterality of the Egg of Chaetopterus 401
SUMNER, F. B., AND P. DOUDOROFF
Some Experiments upon Temperature Acclimatization and Respiratory Metabolism in Fishes 403
STILES, KARL A.
Intermediate-winged Aphids and the Time-of-Determination Theory 430
GALTSOFF, PAUL S.
Physiology of Reproduction of Ostrea virginica, I 461
i INDEX. 487
MBL WHOI LIBRARY
UH 17IS H
* ft
Kl
.
8
131
o
.
,
•
.
m .
DUfJI
•
•
•• »'
.
.
.
.
•
m
Y'A'KV
VOCft
-.'.Vj/
Y.V
'iV.'i
I i OIL
.fill!
I I i I III
II I Cr-
v!' «
. i «r i ill
,v,v
I'.'.V,.'
I I It I • III
III
• i i «
i i i r i i «
i i • i i « i I
it i ..i « f i
C1"'-
ViV
i
i
i i
i
M
.
Bqil
V
ii i
if
O
I*!
It
*
•
I
if
H0KH
•
'llrfj
m
m
ft t
llMI
Wm
r ,JU*j:
•*Mf ''^^B •'
.
-
• .. •
1
I.
i!
i
•
Wv
't
mm m
.
^^^H
I
i
'
t
I
I
.
•
vwt
I