peat
or
a
Yn asia
unin
Hs
bids
ee
, Se. ML the) hh fen) ha aa pare
Digitized by the Internet Archive
in 2009 with funding from
University of Toronto
http://www.archive.org/details/journalofexperim11broo
THE JOURNAL
OF
EXPERIMENTAL ZOOLOGY
EDITED BY
WILLIAM E. Caste FRANK R, LILurr
Harvard University University of Chicago
Epwin G. CoNKLIN Jacques LorB
Princeton University Rockefeller Institute
Cuar_es B. DAvENPORT Tuomas H. Moraan
Carnegie Institution Columbia University
Horacr JAyYNe& GrorGcEe H. Parker
The Wistar Institute Harvard University
HERBERT 8S. JENNINGS Epmunp B. WItson,
Johns Hopkins University Columbia University
and
Ross G. HARRISON, Yale University
Managing Editor
VOLUME 11
1911
‘
THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY
PHILADELPHIA, PA.
call Gp
ye
Rhone a ee y , 7
A RY 4 \ \ Mi, Pad rs
4 : an
me Zs yd it.
18% CoN AE &
We NSS Ae iene ify
shee Lif ae |
x ey Osi
NEN
CONTENTS
1911
No.1. JULY 5
H.S. Jennrnes. Assortative mating, variability and inheritance of size, in
the conjugation of Paramecium. Sixteen figures..........
LoranpE L. Wooprurr AND GuorGe A. Baritsetit. The reproduction of
Paraemecium aurelia ina ‘constant’ culture medium of beef extract.
TINS liVGUROS <0 claheo bp GEER TOR RIO OceOo Goon ae ae bce ares
No. 2. AUGUST 20
Ruty B. Howranp. Migration of retinal pigment in the eyes of Branchipus
eli Us eee OUI PUITES He yore ere feuroyereieis chat erie eras acre te Tere ERE crete ols test «
Davenport Hooker. The development and function of voluntary and
cardiac muscle in embryos without nerves. Fifteen figures
No. 3. OCTOBER 5
C. M. Cuiip. Studies on the dynamics of morphogenesis and inheritance in
experimental reproduction. II. Physiological dominance of anterior
over posterior regions in the regulation of Planaria dorotocephala.
ae tksyy
NEMO TTS. oe od oo ORE nA Motos ord Goes Neca Tear meee
C. M. Cutty. Studies on the dynamics of morphogenesis and inheritance in
experimental reproduction. III. The formation of new zodids in Plana-
Rid eaNGOUNerstOnmMs smn y=SiKd PULES= joe (efota cele a's of-:s «= \cie\« alalelaaleeemies ‘
H. V. Witson. On the behavior of the dissociated cells in Hydroids, Aleyo-
ATI AC AN CMASLOLIAA MMC DIT byt PUTER cele se). sisicicis <),ec0'~ oiels phnietnelalelateiebnrets 2
ili
143
159
lV CONTENTS
No. 4. NOVEMBER 20
LoranpvE Loss Wooprurr AND GEORGE ALFRED BarrsELL. Rhythms in the
reproductive activity of Infusoria. Thirteen charts.................... 339
G. H. Parker anp H. M. Parsuiry. The reactions of earthworms to dry and
to. moist: surfaces .aigieaerenPe renee mies. siaecieciclcdeehee eee eae 361
T. H.Morcan. An attempt to analyze the constitution of the chromosomes
on the basis of sex-limited inheritance in Drosophila. One colored plate
Gour: figures) ia 7sepeepreen cee oer eaoe ocr eer Ae ee Oe 365
E. J. Lunn. On the structure, physiology and use of photogenic organs,
with special reference to the Lampyridae. Nine figures................. 415
Stewart Paton. Experiments on developing chicken’s eggs..:.............. 469
INDEX
Arenas and Asterias, On the behavior of
the dissociated cells in hydroids, 281.
Asterias, On the behavior of the dissociated cells in
hydroids, Aleyonaria and, 281.
Base: G. A., L. L. Wooprurr and. The
reproduction of Paramaecium aurelia in a
‘constan:’ culture medium of beef extract, 135.
Baitset, G. A., L.L.Wooprurrand. Rhythmsin
the reproductive activity of Infusoria, 339.
Beef extract, The reproduction of Paramaecium aure-
lia in a ‘constant’ culture medium of, 135.
Branchipus gelidus, Migration of retinal pigment in
the eyes of, 143.
Csr muscle in embryos without nerves, The
development and function of voluntary and,
159.
Cells in hydroids, Aleyonaria and Asterias, On the
behavior of the dissociated, 281.
Chickens’ eggs, Experiments on developing, 469.
Cuixp, C. M. Studies on the d¥namics of morpho-
genesis and inheritance in experimental repro-
duction. IJ. Physiological dominance of an-
terior over posterior regions in the regulation
of Planaria dorotocephala, 187.
III. The formation of new zodids in Planaria
and other forms, 221.
Chromosomes on the basis of sex-limited inheritance
in Drosophila, An attempt to analyse the con-
stitution of the, 365.
Conjugation of Paramecium, Assortative mating,
variability and inheritance of size, in the, 1.
Culture medium of beef extract, The reproduc ion
of Paramaecium aurelia in a ‘constant,’ 135.
OMINANCE of anterior over posterior regions in
the regulation of Planaria dorotocephala. II.
Physiological, 187.
Drosophila, An attempt to analyse the constitution
of the chromosomes on the basis of sex-limited
inheritance in, 365.
Dynamics of morphogenesis and inheritance in
experimental reproduction, Studies on the,
187, 221.
Vv
Be worss to dry and to moist surfaces, The
reactions of, 361.
Eggs, Experiments on developing chickens, 469.
Eyes of Branchipus gelidus, Migration of retinal
pigment in the, 143.
Hoes: Davenport. The development and
function of voluntary and cardiac muscle in
embryos without nerves, 159.
Howranp, Ruta B. Migration of retinal pigment
in the eyes of Branchipus gelidus, 143.
Hydroids, Aleyonaria and Asterias, On the behavior
of the dissociated cells in, 281.
nFusoriA, Rhythms in the reproductive activity
of, 339.
Inheritance in Drosophila, An attempt to analyse
the constitution of the chromosomes on the
basis of sex-limited, 365.
Inheritance in experimental reproduction, Studies
on the dynamies of morphogenesis and, 187,
221.
Inheritance of size in the conjugation of Parame-
cium, Assortative mating, variability and, 1.
Enns. H. S. Assortative mating, variability
and inheritance of size, in the conjugation of
Paramecium, 1.
AMPYRIDAE, On the structure, physiology and
use of photogenic organs, with special reference
to the, 415.
Lunp, E. J. On the structure, physiology and use
of photogenic organs, with special reference to
the Lampyridae, 415.
Miceatios of retinal pigment in the eyes of
Branchipus gelidus, 143.
Morean, T. H. An attempt to analyse the consti-
tution of the chromosomes on the basis of
sex-limited inheritance in Drosophila, 365.
vl INDEX
Morphogenesis and inheritance in experimental
reproduction, Studies on the dynamics of,
187, 221.
Muscle in embryos without nerves, The development
and function of voluntary and cardiac, 159.
Peeectem, Assortative mating, variability and
inheritance of size in the conjugation of, 1.
Paramaecium aurelia in a ‘constant’ culture medium
of beef extract, The reproduction of, 135.
Parker, G. H. and Parsutey, H. M. The reac-
tions of earthworms to dry and to moist sur-
faces, 361.
ParsHuey, H. M., Parker, G. H. and. The reac-
tions of earthworms to dry and to moist sur-
faces, 361.
Parton, Stewart. Experiments on developing
chickens’ eggs, 469.
Photogenic organs, with special reference to the
Lampyridae, On the structure, physiology
and use of, 415.
Planaria and other forms, III. The formation of
new zodids in, 221.
Planaria dorotocephala, IT. Physiological domin-
ance of anterior over posterior regions in the
regulation of, 187.
EGULATION of Planaria dorotocephala, IT. Physi-
ological dominance of anterior over posterior
regions in the, 187.
Reproduction of Paramecium aurelia in a ‘con-
stant’ culture medium of beef extract, 135.
Reproduction, Studies on the dynamics of morpho-
genesis and inheritance in, experimental, 187,
221.
Reproductive activity of Infusoria, rhythms in the,
339.
Retinal pigment in the eyes of Branchipus gelidus,
migration of, 143.
Rhythms in the reproductive activity of Infusoria,
339.
Seeuncren inheritance in Drosophila, An at-
tempt to analyse the constitution of the chro-
mosomes on the basis of, 365.
VV orentarr and cardiac muscle in embryos without
nerves, The development and function of,
159.
Wo; H. V., On the behavior of the dissoct-
ated cells in hydroids, Aleyonaria and
Asterias, 281.
Wooprurr, L.L. and G. A. Barrsett. Rhythms
in the reproductive activity of Tnfusoria, 339.
Wooprtrr, L. L. and G. A. Bartsety. The repro-
duction of Paraemecium aurelia in a ‘constant’
culture medium of beef extract, 135.
Zoers in Planaria and other forms. III. The
formation of new, 221.
ASSORTATIVE MATING, VARIABILITY AND INHERIT-
ANCE OF SIZE, IN THE CONJUGATION OF
PARAMECIUM'
H. 8. JENNINGS
From the Zoélogical Laboratory, Johns Hopkins University
SIXTEEN FIGURES
CONTENTS
Problemsioutlimed ey. cresctecers rare erste rete eter eretetolelole Cieverolaistate re aheleteielsle\s @ eie' ele 3
I. The facts as to relative size, variability, and assortative mating in con-
UGH) Vise Soc chaeesoqueueEDpouseLocchboadicodeubr-acsoo Coo UpoOOOOIaE 6
MIGHNOOE: <3 .aocdaabdGapSCsoagooen OE snOGo oduUSEUneOScd dd Cguect.oc. OsnnaeMEE 7
Hundamentallmeasurementss)). 5 0... sic1c cc a tnseu s+ siecle cee cieleinie sie ssie we 9
Relative size of conjugants and non-conjugants...........-.-6+++52seseeee 11
: Relative variability of conjugants and non-conjugants.........-.....5--+5 16
Causes of the lessened variability of conjugants ............+..-20.500005 17
iGametredifferentiations assets eect toot teet retire Astle = «= 17
PA JOG EIIt cholo lated iach GoDoriO DoGHnoOd nodes Saee aU cod do LoD aE eE gor 18
Bi (Condo cocceoncoictobo paodo ond cee rEn en toon.caac scp soso.00 Cnoeoeee ols
Increase in size and ceuability of the conjugants before fission. .... 20
Am IVACtAINCITErENCeS he nes cee cease ech eias sia ase soi cietncrns as * os + 22
ANOUK DIC WR op COA SeS GaGa RG sonnet Come on oon aodee eoneac Onn iene 23
ReaniisvexplanahiOMerre tem citrc pei cis erect ststevate stleeietereia stsia. =o) oe =! 23
Observation of the process of conjugation............-.-.-.eeeeeeee rere: 24
Correlation in size in the members of pairs of conjugants.............--- 2 228:
Meaning of the coefficients of correlation.....:........++eeeee eee eeeeeetee 28
Correlation in different classes of cultures...............00eeee eee ee eee SH
Wild cultures, of unknown racial composition..............5...05++ 5555 37
Cultures containing pairs belonging to two different species..........-- 39
Conjugation within pure races... 2. ..2c.22sscccc eet eres snes eens 41
Mixturestof tworknoOwn-Tacess.-... 2-2-2200 s0s+s RAR Reno pCoC 44
To what is due the incompleteness of correlation?...........---++++s55055 46
1 Slight differences of less effect than great ones..........-----+-+55050 46
2 Different categories of pairs following different rules........-..+.+-- 51
Unevenness at the anterior ends.............. MRP teased o 8 nth Cac 52
1 Third of a series of papers on Heredity, Variation and Evolution in Protozoa.
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, No. 1
Jury, 1911
il
2 H. S. JENNINGS
Causesvof the; correlation: {04-4 Meee seen: 3.) «oe oe «ae 60
Lg pAssortative mating... cians an eos, eee. « = eee s (a0)
a Correlation between the length anterior to the mouth and the
total length 5.4 ae eet ects conte Ce ee ee Oe 61
b High variability of the posterior region........................ 63
ce Low correlation of parts before and behind mouth.............. 64
d Correlation of pairs greater for parts anterior to mouth ........ 64
e Low correlation of parts posterior to mouth.................... 64
f Region in front of the point where the two conjugants separate... 65
2 Equalization during mating............. taGinged eon ee Bs as
Correlation of pairs after eepamion POEM He eM oe os Aap Stic. c 67
Change in variability and correlation of parts after separation ..... 70
Correlation of pairs not due to equalization....................... 73
3 \Changelofisizeduning union’... sete: coe ceo eee eee 7:
4 Differential contraction due to killing fluid......................... 77
5 Local or temporal differentiations in the culture.................... 80
6) Othersuggested causes a2 cc. -..c.2e Ree en eee oe noe Renee 82
@ (@orrelationsin /breadthe.......s.cs.ceseeeee PR ODA SrA c 82
a Flattening at the time of conjugation........................02- 83
b Correlation in breadth in pairs after separation................ 84
ec Correlation in breadth not due to equalization.................. 85
Siebostoricalyandtcomparatlv.er sane eee eee ee Eee 85
Se Conclusionsiontassortativie:matim ge...) veeee aera eric eee ea reielo sia 89
II Consequences of the differentiation of the conjugants and of their assort-
ative-mabin pieces ore en eet Eee eae oer 39
Consequences of the decreased size and lessened Tariab ity SARS aD aac turne 90
1 Are the extreme specimens excluded from the new generation?........ 90
2 Relative size of progeny of conjugants and non-conjugants; changes in
size due to conjugation.......... FS ANS | RPE eas 92
a Comparison of progeny of conjugants and non-conjugants in pure
0: "=< RE fo an ey a te cs ae iaeio LG eo toe shoneleHs Buccs eee merged 93
b Comparison of progeny of conjugants and non-conjugants in a
wildvcultures. 28.0 tc. Mae nes eenas ecnetet ate os vensters onteses:cesbareereecrs 97
NED GLK ald: Grenier. nao ORE Tan Ratton Goma cer acta nator 100
Cabixceptionalicase ac. nmecnia nin erie tie ase cee er 101
CG ASUMIMAL Ys. -Spem tee ic rice Sree Te ere eel il Poe eae ena eaters Cleese: 103
3 Increase in variability as a result of conjugation.................. .. 104
4 Do pairs of different size give progeny of different size? .............. 104
5, Unkeritancesiromiunequal pairs. see ateee meets eit: = 106
Summary, eaysereee a Naa ROe Teor Or aac Un GUb aS dpobodjen coda auemeans cc 106
Literature ‘cited sete rise vcce cet ee se i337 ee 109
110
Appendix: tables of eanrerienia Ee So fC Ads COUCUC ORE IEC Oo o
CONJUGATION IN PARAMECIUM 3
INDEX OF TABLES
LE hee hot ete arene sactsuats re atare 9 1 he ere Pei oko ot ie en ree 62
7 c.g SED PIDICES CRO DCE rae 10 19S EPR peers soe ate ss 69
SMe ne nre ere secrefeia erator 11 PAD es 55 ac gr DRA Onn 72
Caley Rear iG BRR DE EKO oS Corer: 12 he cv SST Bakes 73
Seen gee Ontos avin sche Ceeeetd 12 PRE Ee rho S's RROEe 75
Gs ete cticeistctny Stonsrarny Seapets: eos sire 17 D8 sow asg ae ee ete 76
Ua PEERS Tech oeee erate ome alain 2 toate tes 21 yA. EE Grito aD SOCK 76
IE Para aod an Dosh Gon naom nee 32 25)... i. 9G anes eee 7
(2) An ene, ltr ho et ete eee ee ary eT 36 Ps hams doCo dae 84
(Oitint o dace ombo toanor om com een 40 QE sss svatavgatee ES Ceres 84
1b a teed Coa benhocAconnaonone 47 28). os cil. See ee 94
ID eee sme srs ties crcnccerete ke tepercters 50 7 RREPAGAT nS hic 505 00S 95
NSP Sey men rae peel. Reacanaveyovarete oie 54 BO wc saenick-s Scares ee ee 96
VAN eee is) pepsi) sya iskrie Swear ein Ae 55 0) Pe ee co oct 97
hk tar Aen AaB an Aeterna 56 Sores os ces aistotente-os aoe 99
1G Lays Recodo bckciae ream 57 DO ele aaa ahuadta ares ats, coronas 102
Le, RPI E eee it oe ns 59 34-68, Appendix.......... 110
PROBLEMS OUTLINED
The second paper of this series dealt with variation and inherit-
ance of size in Paramecium during reproduction by fission.
The present paper is an experimental and observational study
of the size relations in conjugation. A later contribution will
deal with the relation of conjugation to vitality and reproduction.
Results presented with biometrical analysis have unfortunately
come to incur in many quarters the suspicion that mathemati-
cal treatment has been substituted for accurate experimental
and observational investigation. The numerical analysis of
results should of course be an addition, not a substitution; but
the experimenter should realize that without this addition ex-
perimental results may sometimes be as misleading as statistics
without experimentation (“which is putting it strong’). Adequate
experimentation with adequate numerical analysis is the ideal;
toward this ideal my efforts, however short they may fall, have
been directed in the present investigation.
In taking up the relation of conjugation to genetic problems,
I have thought it best to become acquainted at first hand, so
4 H. S. JENNINGS
far as possible, not only with matters that have not been hitherto
treated, but even with those that have been dealt with by previous
investigators, in the latter case confirming or criticizing their
results. This does not imply a precedent suspicion as to the
accuracy of the work thus gone over; it is done only in pursuance
of a general policy, for one often finds matters of great import
where they are least expected. Furthermore, the recent dis-
covery of the existence of many diverse races in Paramecium
makes it needful to reéxamine many phenomena in relation to the
part played in them by these different races. In any case, in
this difficult field independent confirmation of another’s results is
decidedly worth while.
It will be well to set forth here an outline of the questions
with which a thorough investigation of the size relations in con-
jugation would have to deal.
To Pearl (07) we owe the discovery of certain most interest-
ing relations between the conjugating individuals of Paramecium.
By an elaborate statistical investigation he showed (1) that the
conjugants of a culture of Paramecium are much less variable
than the non-conjugating population, and have (as had before
been noticed) a smaller mean size; (2) that there is a marked
degree of correlation in size between the members of pairs in Para-
mecium; smaller individuals being found mated with smaller,
larger with larger. With these important matters, particularly
in their relation to the existence of diverse races, we shall have
to deal thoroughly. The precise questions here are as follows:
1. What are the facts as to the relative variability and size
of conjugants and non-conjugants, and what is their relation to
the existence of races of diverse size?
2. What are the facts as to assortative mating; its determin-
ing conditions, its peculiarities and limitations; its relation to
the existence of diverse races?
3. What are the results, in inheritance, variability and evo-
lution, of the smaller size and decreased variability of the conju-
gants, as compared with the non-conjugants? If we breed from
a number of the conjugants, do they give progeny that are (a)
smaller, or (b) less variable, than the progeny of the larger, more
CONJUGATION IN PARAMECIUM oO
variable non-conjugants? Does conjugation thus act as a pro-
cess of excluding from the line of evolution individuals that vary
from the usual size?
4. What are the results, in inheritance, variability and evolu-
tion, of the assortative mating of Paramecium? If we isolate
(a) large, and (b) small, pairs of conjugants, keeping them under
identical conditions, do they produce, respectively, large and
small races? Is there any difference between the progeny of
(a) pairs in which the two members are equal, and (b) pairs in
which the two members are unequal?
5. What relation has conjugation as a physiological process
to the size of the individuals of the stock undergoing it? Does
the size differ characteristically in different parts of the life cycle,
as is sometimes set forth? Are the individuals at the ead of the
life cycle (just before conjugation) larger or smaller than those
at the beginning of the cycle (just after conjugation)?
6. What are the facts of inheritance in conjugation? If the
two conjugants of a pair differ, are the progeny of these two
conjugants alike and intermediate between the two? Or will
for example the larger member continue to produce large individ-
uals like itself, the small one small individuals like itself? Or is
there some third possibility? The laws of inheritance have never
been worked out for this peculiar reciprocal fertilization, where
both parents may continue reproduction.
These questions we shall take up in detail. On most of them
I hope to present data of importance, though on the extremely
important problem last raised I have as yet been unable to get
clear results on some of the points of greatest interest.
In dealing with most of these questions, the existence of diverse
races of Paramecia, as set forth in former papers,? will be found
of extreme importance. In Paramecia multiplying by fission
there are many races or lines, differing in size and in other respects.
A considerable number of these races were isolated; the mean
length of the largest being more than double that of the smallest,
with many intermediate races. As will be recalled, each race
2 See Jennings ’08, and Jennings and Hargitt ’10.
6 H. S. JENNINGS
shows within itself many variations, due to differences in growth
and environmental action, but these variations within the race
are notas arule inherited, and under the same conditions of growth
and environment the race is uniform and constant. In all work
with conjugation, the question whether we are dealing with a
pure race or with a mixture of races is of the greatest importance;
the significance of the phenomena observed is quite different
in the two cases.
I. THE FACTS AS TO RELATIVE SIZE, VARIABILITY AND ASSORTA-
TIVE MATING IN CONJUGATION
The data given by Pearl (’07) would seem amply sufficient to
show that in a conjugating culture the conjugants are smaller
and less variable than the non-conjugant population, and that
there is a high degree of assortative mating in Paramecium. A
further study of the facts is needed, however. On the one hand
it is desirable that Pearl’s interesting results should be confirmed
by independent observation, or refuted. The correctness of
some of his results has been called in question (Lister ’06, Pear-
son ’06, Pearl ’07). Further, there are a number of conditions
not dealt with by Pearl that might produce a correlation in size
between the members of pairs; these need to be subjected to
experimental test. Beyond this, many important relations in
this matter remain as yet unknown; we need a knowledge of the
variations and limitations of assortative mating, of the conditions
on which it depends, of its relation to the existence of diverse
races, and particularly of its consequences in the later history of
the stock.
I have therefore examined and made measurements of a num-
ber of conjugating cultures with reference to these matters. After
a number of cultures had thus been studied, it became necessary
to test by comparative examination of cultures under controlled
conditions, one after another, many factors that suggested them-
selves as possibly producing the observed relations (particularly
the correlation between members of pairs). Asa result, the quan-
tity of material available for study of these matters becomes
CONJUGATION IN PARAMECIUM Ui
very great, consisting of more than thirty lots, averaging more
than one hundred pairseach. Some of these lots were ‘wild’
cultures, containing a number of diverse lines or races, belonging
in some cases all to caudatum or all to aurelia; in other cases
belonging partly to caudatum, partly to aurelia. Other lots
consisted entirely of members of a single race or ‘pure line,’
having descended from a single individual; other lots consisted
of mixtures of two known races. The relations observed are
naturally somewhat different in these diverse cases. I have
dealt mainly with the measurements of length, since it is here
that the phenomena of primary importance appear; certain stud-
ies of the breadth relations will however be found on later pages
(table 20).
METHODS
The special methods for the diverse experiments will be men-
tioned in the course of the paper. Here it will be well to mention
mainly the methods of killing and of measurement.
The animals to be measured were brought into a drop of water
at the bottom of a watch-glass, then overwhelmed with the kill-
ing fluid. For killing I used mainly Worcester’s fluid (10 per cent
formalin saturated with corrosive sublimate). I have later found
that chrom-osmic acid (1 per cent osmic in | per cent chromic)
has advantages in some respects. Both these fluids kill without
distortion if properly used. The animals were measured either
in the killing fluid, or after transference to water or to 25 per
cent glycerine. Careful comparative measurements before and
after transference showed that no change is made by placing in
water or weak glycerine. The most satisfactory method is to
remove with pipette a portion of the killing fluid from the watch
glass, then to add 25 per cent glycerine; in this the specimens
are kept till measured. With the Worcester’s fluid there is some-_
times an objectionable deposit of fine crystals, in the course of
time; this does not happen with the chrom-osmic.
In the early part of the work the animals were measured on the
slide, with the ocular micrometer. This becomes very wearing
8 H. S. JENNINGS
on the eyes; later the measurements were made by the aid of the
Edinger drawing and projection apparatus, which cannot be too
highly recommended for the purpose. The animals are projected
much enlarged, on the drawing board, where they are measured
with a millimeter ruler. I used a magnification of 500 diameters,
so that each millimeter of the ruler corresponded to 2 microns
(0.002 mm.). The best method I found to be as follows: the ani-
mals were placed on a thin slide in a flat drop of the 25 per cent
glycerine, with no cover (so that there was no danger of flatten-
ing), projected, and measured. Without the glycerine in the
fluid this method cannot be used, owing to the convection cur-
rents and the rapid evaporation produced.
In the measurements of conjugants the unit of grouping was
4 microns, so that each group in the tables contains individuals
varying from 2 microns below to 2 microns above the dimen-
sion at the head of the column or row. In the original measure-
ments, in many cases, the unit employed was but 2 microns.
The constants of variation were computed according to the
methods and formulae set forth in my paper of 1908 (page 397).
In the present paper however we are dealing with cases where
the two things to be compared (the-two members of a pair) are
alike, so that either one might be entered in the rows or in the col-
umns of the correlation table. In such cases double or symmet-
rical tables have commonly been employed. In a recent note
(11) I have shown that this is unnecessary, and that the compu-
tations are performed with much less labor without the use of
symmetrical tables. The method of computation set forth in
this note was used in the present paper. In accord with this, |
have formed the tables of correlation by entering in every case the
larger member of the pair in the vertical columns, the smaller in
the horizontal rows.
CONJUGATION
IN PARAMECIUM
FUNDAMENTAL MEASUREMENTS
Table 1 gives the important constants for the length of con-
jugants as compared with non-conjugants in a number of cul-
tures developed in material brought into the laboratory from
ponds or pools, so that the racial composition is unknown. Table
2 gives the same constants for cultures consisting entirely of a
single ‘pure line’ or race,—all the individuals being derived from
the fission of a single one; also those for certain mixtures of known
Table 3 gives the constants for a number of lots of con-
races.
TABLE 1
Constants of variation in length for conjugants and non-conjugants of Paramectum, from
wild cultures, of unknown racial composition.
(The original measurements of length
for all these will be found in table 34 of the appendix; the tables of correlation for the
conjugants, in the appendiz, are indicated in the column headed ‘table’ )
tw
~
ou
wo
LOT
DATE
Dec. 13,
Mar. 21,
Mar. 26,
Feb. 17,
\Jan. 29, .
Apri) ost
Nov. 4, ’07 |
June 20,’09
07
"08
"08
08
| 1 D 1 fu & i
| 1535 BA zis ill uae eS | fEze
Bl pe ras AS Bra Bg | faae
RBZ q| aoe a ags Sipe Beege
leeles|a|eee| 32 Z88 2 | Seas
all aaa teed cei ae Bee 8 oa
A. ‘‘Wild” cultures | |
a. Conjugants........ | 360 180) 11 148-260199.02+0.54 15.28+0.38 7.68+0.190.39840.030
b. Non-conjugants ..| 360) 34 132-320 222.86+0.87 24.57+0.62 11.03+0. 28
a. Conjugants........ | 284 142) 36 130-206 165.04+0.53) 13.20+0.37, 8.00+0.23/0.268+0 057
b. Non-conjugants.... 87] 34) 134-198 164.12+1.00 13.84+0.71 8.43+0.43)
a. Conjugants........ 164, 82) 37 164-236196.100.71) 13.54+0.50) 6,91+0.260.507+0.049
b. Non-conjugants.., 176 | 34 164-288 228.36+1.23) 24.20+0.87| 10.60+0.39
a. Conjugants ....| 84) 42) 38 116-160/139.29+0.70) 9.56+0.50) 6.86+0.36(0.499+0.055
b. Non-conjugants ...) 152 34 100-244155.40+1.39 25.42+0.98 16.36+0.65
a. Conjugants, de-
scended from 4-a.., 16, 8 34 116-144130.50+1.33) 7.90+0.94 6.05+0.72
b. Non-conjugants
descended from 4-a., 100 34 104-200 143.80+1.29 19.08+0.91 13.27+0. 64
a, Conjugants........ | 272) 136] 39 128-216181.49+0.54) 13.32+0.39 7.34+0 21/0.428=0 033
b. Non-conjugants ...| 318} 34) 132-248)/186.10+0.79| 20.97+0.56) 11.27+0.31)
fa. Conjugants........ | 158| 79| 40) 128-204 168.71-0.63| 11.66+0.44 6.91-+0.26(0.333+0.048
| (b. Non-conjugants ..) 131) | 34) 148-224'182.20+1.03| 17.43+0.73) 9.57+0.40)
| B. Descended from se-| |
lected parts of wild) |
cultures }
2a.Conjugants de-
scended from 10
BM oa nocoods 54) 27) 42 120-152134.89+0.64 7.02+0.46 5.20+0.340.612+0.057
b. Non-conjugants, |
( froin same......... 56 35) 124-168148.93+0.98, 10.89+0.69) 7.31+0.47
10
H. 8S. JENNINGS
TABLE 2
Constants of variation in length for conjugants and non-conjugants of Paramecium, from
cultures of pure races, descended from a single individual or a single pair. or from
mixed cultures of known racial composition.
(Lhe measurements of length for these will
be found in table 25 of the appendix; the tables of correlation for the conjugants, in the
appendix, are indicated in the column headed ‘table’ )
|
Lil 4 ee le ees ;
Jaen Bei | cau eecemulncees
S28 | |Fza] 29 ae |) ae Z2h2
Paar) ra AS Hz wp Be Bean
Helge ,|a5¢| 73 | 382 | ES | SBEee
a 3 SelB) Cees) Ae zee | 28 ESfa
5 < Z papa 2 o<2 ae oi) ay Boman
a a fof a 2 |B| & a nD o is)
A. Pure lines from
| one individual }
| | (all aurelia)
alsapes 28,071 - Conjugants...... 250, 125 43) 120-180 150.50+0.48, 11,130.34, 7.39+0.22) 0.132+0.042
tay ib. Non-conjugants 200 35] 124-200 188.80+0.88 18.38+0 62 11.58+0. 40
10 Feb. 20, 08 z Conjugants...... 52, 2644) 112-144 127.23+0.67, 7.14+0.47 5.61+0.37\—0. 193+0.090
es b. Non-conjugants | 43 —35| 120-160 140.19+0.99 9.65+0.70, 6.89+0.50
|
11 Feb. 26, '08) i k. Conjugants...... | 98) 1445 06-136 116.71+1.13 8.87+0.80 7.60+0.69|—0.137+0.125
Forenoon b. Non-conjugants | 100 35 96-152 133.68+0.79 11.64+0.56 ».70+0.42
12\Sept. 12,'08) & tle Conjugants...... 156 78 46 104-140 124-08+0.41 7.51+0.29 6.05+0.23) 0.367+0.047
Se eae eel b. Non-conjugants | 100, 35) 104-180 143.72+0.96| 14.25+0.68| 9.91+0.48}
| | |
Afternoon k a. Conjugants...... 336 168,47 92-156 129.58+0.40 10.96+0.29) 8.46+0.22) 0.184+0.036
13 Sept. 12, '08) b. Non-conjugants 100 35] 85-168. 140. 20+0.97 14.35+0.68, 10.23+0.49
1alMar. 31, 08 Nfs (/" Conjugants...... 42, 21,35] 124-148 136.95+0.58 5.53+0.41) 4.03+0.30| 0.295+0.095
i piak iw | b. Non-conjugants 34 BE 132-168 144.59+0.92 7.92065 5.48+0.45
islApr. 4 “08 WE a. Conjugants...... | 50 25.25) 116-148 132.88+0.64 6.66+0.45, 5.01+0.34) 0.257+0.089
Paget * \'b. Non-conjugants | 31, —(/35| 120-188 147.61+2.20| 18,181.56) 12.31+1.07
1eSept.14, "08 Cs a. Conjugants...... | 300 150,48} 100-160 128.67+0.47 11.97+0.33 9.30+0.26] 0.507+0.029
ert * |b. Non-conjugants 100 35 96-168 134. 2041.04 15.37+0.73, 11.45+0.55
17Sept. 16°08 Cp {2 Coniugants...... | 188 6949 88-148 121.91+0.66 11.46+0.47 9.40+0.39) 0.318+0.052
AOE eres Me lb. Non-conjugants 110 [35] 104-164.132. 180.87) 13.53+0.62) 10.2340. 47
| | |
Forenoon | If Conjugants...... | 168, 8450) 104-152 123.57+0.41) 7.80+0.29 6.31+0.23) 0.251+0.049
1gSept. 25, '08 7 th. Non-conjugants | 100 [35] 88-180 131.32+1.21) 17.97+0.86 13.680. 67
| | | |
Afternoon | a. Conjugants. ..... | 174 87)51| 96-152 118.28+0.£2) 10.11+0.37, 8.55+0.31! 0.323+0.046
19 Sept. 25, 08 te. Non-conjugants | 118 35) 88-180 135.35+1.02 16.49+0.72) 12.18+0.54
| B. Mixtures of two |
| known races | |
aiieaens 108 C2 4. / 8 Conjugants are 124 6253) 100-156 129.58+0.62 10.31+0.44, 7.95+0.34) 0.115+0.060
la |/b. Non-conjugants | 149/35) 84-176 122. 44+1.37, 24.71+0.97 20.18+0.82
SNe ie ns ta+k [ Conjugants...... 98 4954) 108-136 120.25+0.46 6.68+0.32 5.56+.27| 0.4080. 064
ia jeans Sean b. Non-conjugants. 156 ia 104-264173.10+2.25 41,671.59) 24.07+0.97
jugants where the corresponding non-conjugants were not
examined. ‘The original measurements on which these constants
are based will be found in the tables of the appendix; the more
important ones in tables 34 and 35.
CONJUGATION IN PARAMECIUM J]
TABLE 3
Constants of variation in length for a number of lots of conjugating Paramecia in which
the non-conjugants were not measured. The column headed ‘table’ gives the number of a
table to be found in the appendix, in which the distribution of the measurements is
shown. The measurements are here given in microns
, 4 he & fa fe
z Fe E 3 OF EO
pb = 5 z z =
Satemlonlt gj ag go Rake
jog B| ee é 4 me Bie Baga
| ja alag| . ag << Ss | 88pea
Q asians) & On z a Be Beads
g| & Balpa|&| 3 F 23 Bs | aSaza
3 | A 2 \% | a| # 2 a 8 8
|
| | | |
| 1910 | |
22 Aug. 31 Wild culture, caudatum) 204) 102 55 | 152-208176.14+0.48 10.08+0.34 5.72+0.190.359+0.041
| | 22
| | | ing
23} Sept. 21 ‘Wild, caudatum ...... 296) 148.and) 152-208179.80+0.41 10.42+0.29 5 80+0.160.245+0.037
| | 23 |
24 Sept. 11 |/Racek (aurelia)........ | 244 122 59 100-144.118.92+0.33, 7.54+0.23 6.34+0.150.210+0.041
26 Sept. 11 Mixed caudatum ca |
erauitvel tals rereiyeerareretereserete | 62) 31 84-200145.81+2.44 28 491.73 19.54+1.230.939+0.010
27 Sept. 13 Mixed caudatum and) |
BUTELIAUS ascs scheint 340, 170 10 108-236161.19+1.27 34.70+0.84 21.53+0.540.940+0.004
With certain exceptions, to be considered later, the results given
in these tables confirm Pearl’s results (1) that the conjugants are
smaller than the non-conjugant population of a culture; (2) that
they are less variable than the non-conjugants, and (3) that there
is a marked positive correlation in size between the members of
the pairs, so that on the whole larger individuals are found mated
with larger, smaller individuals with smaller. We shall take up
these matters separately.
RELATIVE SIZE OF CONJUGANTS AND NON-CONJUGANTS
If we examine in the seven ‘wild’ cultures of table 1 the relative
mean lengths of the conjugants and non-conjugants of a culture,
we find, as shown in table 4, that in every case save one the con-
jugants are smaller than the non-conjugants. In lots 1, 4 and 5,
the difference between the means for the conjugants and non-
conjugants is about 10 per cent of the mean length of the non-
conjugant population; in lot 3 it is 14 per cent. But in lot 6
the difference is shght, being but 2.5 per cent of the mean for the
12 H. S. JENNINGS
non-conjugants, and in lot 2 the mean length of the conjugants
is actually the greater, by a very small amount, though here the
difference is not significant in comparison with the probable error.
In this culture then the conjugants are not perceptibly differ-
entiated in size from the non-conjugants.
In the four lots studied by Pearl (07) the conjugants were in
all cases smaller than the non-conjugants, by amounts varying
from 11.5 per cent to 16.4 per cent of the mean of the latter, and
TABLE +4
Differences in length between conjugants and non-conjugants of wild cultures. (The
‘relative difference’ in the fourth column shows what percentage the difference is of
the non-conjugant mean)
NON-CONJUGANT
LOT MEAN CONJUGANT MEAN ABSOLUTE DIFFERENCE RELATIVE DIFFERENCE
Per cent
1 | 222.86+0.87 199.02+0.54 23.83+1.02 10.7
2 164.12+1.00 165.04+0.53 | —0.93+1.13 —0.54
3) 228.36+1.23 196.10+0.71 | 32.27+1.42 | 14.1
4 155.40+1.39 139.29+0.70 | 16.11+1.56 | 10.4
5 | 143.80+1.29 130.50+1.33 | 13.30+1.85 9.2
6 | 186.10+0.79 181.49+0.55 | 4.61+0.96 2.5
7 | 7.4
182.20+1.03 168.71+0.63
13.49+1.20
TABLE 5
Differences in length between conjugants and non-conjugants of pure races and
mixtures of pure races
Per cent
9 c 158.80+0.88 150.50+0.48 8.30+1.00 5.2
10 k 140.19+0.99 | 127.23+0.67 12.95+1.20 9.2
11 k 133.68+0.79 | 116.71+1.13 | 15.97+1.38 11.9
12 k 143.72+0.96 124.08+0.41 19.64+104 | 13347
13 k 140.20+0.97 | 129.58+0.40 | 10.62+1.05 7.6
14 Nfo 144.59+0.92 136.95+0.58| 7.64+1.08 5.3
15 Nf 147.61+2.20 132.88+0.64 | 14.73+2.29 10.0
16 C2 134.20+1.04 128.67+0.47) 5.5341.14 Ae
17 C2 132.18+0.87 121.91+0.66 10.27+1.09 WES
18 g 131.32+1.21 123.57+0.41 7.75+1.28 5.9
19 g 135.35+1.02 118.28+0.52) 17.08+1.15 12.6
20 Co4t 122.44+1.37 129.58+0.62 —7.14+1.50 —5r8
30.6
21 Lak 173.10+2.25 120.25+0.46 52.86+2.30
CONJUGATION IN PARAMECIUM 13
it has been practically the universal testimony of observers that
the conjugants are smaller than those not conjugating. Our
own results, as we have seen, confirm this for most cases, but not
for all. How is the fact to be accounted for that in some cultures
the conjugants are not smaller?
Light on this question will best be obtained by examining the
relative sizes of conjugants and non-conjugants in cultures com-
posed of pure races, and in mixtures of known racial composition.
The data are given in table 2 and in table 5. In all the eleven
cases of table 2 in which we can compare the conjugants and non-
conjugants of a pure race, we find the conjugants smaller, by
amounts varying from about 4 per cent up to more than 12 per
cent, of the mean for the non-conjugant population. All of
these are races of aurelia, as I had no opportunity to make a
careful study of the conjugants of a pure race of caudatum. But
the fact that in wild cultures consisting mainly if not entirely of
caudatum, as was the case with all of Pearl’s material, and of
our lots 1,3, 4and 5, the conjugants are as a rule markedly smaller
than the non-conjugants, indicates strongly that this would
hold generally for caudatum also. We may then take it as
established for aurelia, and practically so for caudatum, that
within any given race the conjugants average smaller than the non-
conjugants. Why in some wild cultures the conjugants may not
be found smaller is shown by examination of the data for our
mixed cultures (lots 20 and 21, table 2). Lot 20 consisted of a
mixture of two races of aurelia,i and (>. The race 7 was smaller,
averaging usually about 100 microns in length, while the usual
mean for C, was about 125 microns. When conjugation took
place in this mixture, the conjugants were all of the size charac-
teristic for the conjugants of C, (as shown by comparing lots 16,
a and 20, a of table 2), measuring 129.58 microns. Conjugants
in a pure race of 7 had been found to be much smaller, varying
from 92 to 98 microns in length. Thus it was clear that in the
mixture only the race C, was conjugating, and the measurements
1 For measurements of these races uader various conditions, see Jennings ’08,
and Jennings and Hargitt 710.
14 H. S. JENNINGS
for the conjugants are of that race alone. But the random sample
of non-conjugant population contains representatives of both 7
and C2, and its mean size (122.44 microns), therefore lies between
that of the two races. It is therefore less than that of the con-
jJugants. The conditions in this case are illustrated in fig. 1.
They are well brought also by acomparison of the measurements of
the conjugants and non-conjugants of lot 20, as given in table 35.
Al
JIN
/ i
Fig. 1 Typical group of specimens from a culture consisting of a mixture of
the large race C2 and the small race 7 (both aurelia). Conjugating pairs all (>.
X 333.
The reverse condition is given by the mixture of LZ. and k
(lot 21, tables 2,5 and 34). Here the conjugation was only in the
smaller race k, while the non-conjugant sample includes also many
of the large race Z2, As a result the conjugants are very much
smaller than the mean for the mixture as a whole, the difference
being 30.6 per cent this mean.
CONJUGATION IN PARAMECIUM 15
These cases show that when more than one race is present in
a mixture, the members of one race alone may conjugate. If this
is a large race, the mean size of the conjugants may be equal to
that of the population as a whole, or even larger than this. This
is doubtless the explanation for lots 2 and 6, tables 1 and 4; here
we are dealing with cultures of unknown racial composition. As
a matter of fact I did, by selection and propagation, isolate a
number of races of diverse size from lot 6; from it came the races
k and Ls, as well as a number of others.
Thus when we are dealing with a single race the conjugants
are always smaller than the non-conjugant population, by amounts
varying from 4 per cent to 13 per cent (or more) of the mean
for the latter. The variation in the proportional difference be-
tween the two in different cases is readily accounted for the fact
that sometimes multiplication is occurring during an epidemic
of conjugation, at other times not. In the former case many
young will be present, reducing the mean length for the non-
conjugants, but not affecting that for the conjugants. It is to
be noted however that the mean size of the conjugants may differ
a certain amount under different conditions in the same race. In
the race k the mean size at the epidemic of February 26, 1908,
was 116.71 microns (table 2, lot 11) while at the epidemic of Sep-
tember 12, 1908, it was, in the afternoon, 129.58 microns (lot 13),
a difference of 11.02 per cent of the smaller mean size. This is
the greatest difference in mean size observed between the con-
jugants of a given race. It is to be observed that this difference
is not one arising progressively over long periods, for but a week
before the minimum, the conjugants of this same race showed
practically the maximum size (compare lots 10 and 11, table 2).
The difference is undoubtedly due to the varying nutritive condi-
tions at the time of conjugation.
But in cultures of unknown racial composition, the conjugants
may be very much smaller than the average for all (as in lot 21),
or they may be equal to or larger than the average for all, depend-
ing on the relative size of the races present, and upon which of
these races the conjugants belong to.
16 H. S. JENNINGS
The consequences in heredity of the decreased size of the con-
jugants will be taken up later.
RELATIVE VARIABILITY OF CONJUGANTS AND NON-CONJUGANTS
Examination of tables 1 and 2 confirms further Pearl’s dis-
covery that the conjugants are not only smaller, but also less
variable than the non-conjugant population. A comparison of
the original measurements for the conjugants and non-conju-
gants, as given in tables 34 and 35 of the appendix, renders this
difference in variability at once evident to the eye. In every one
of the cases given by these tables, both the absolute variation
(as shown by the standard deviation) and the relative variation
(as shown by the coefficient of variation) are less in the conju-
gants. In a few cases the difference is but slight and would
perhaps be hardly significant, taking each of these cases sepa-
rately, in comparison with the probable errors. But the fact
that it is always the conjugants which show the lesser variability
is very significant, especially when we consider the much more
numerous cases in which the variability of the conjugants is much
less than that of the non-conjugants.
The difference between the variability of conjugants and non-
conjugants itself varies greatly in different cases; in other words,
sometimes the conjugants are but little less variable than the
non-conjugants, while in other cases they are very much less vari-
able. This is exhibited in table 6. In some of the wild cultures
the coefficient of variation of the conjugants is but 5.1 per cent
less than that of the non-conjugants (lot 2, where the difference is
indeed of no significance in comparison with the probable error) ;
from this minimum it varies uv to a difference in lot 4 of 58.1
per cent,—the coefficient of variation for the conjugants being
less than half that for the non-conjugants. In the pure races
the least difference between the coefficients for the conjugants
aod non-conjugants is 8.1 per cent of that for the non-conjugants
(lot 17), rising to 59.3 per cent in lot 15. If we make an average
of these numbers showing the difference in variability for the eleven
lots of these races in table 6, we find it to be 33.01 per cent.
CONJUGATION IN PARAMECIUM 17
CAUSES OF THE LESSENED VARIABILITY OF CONJUGANTS
What is the cause of this lessened variation among the conju-
gants? A uumber of different possible factors may be considered.
1. Gametic differentiation
Lister (06) suggested that the cause was as follows: The
conjugants are differentiated gametes. In measuring them we
are dealing only with this particular class, while in the non-con-
jugant population we include many gametes, many in the process
of differentiation into gametes and many that are not gametes;
hence the non-conjugants are a heterogeneous lot and must give
TABLE 6
Difference in variability between conjugants and non-conjugants
STANDARD DEVIATION COEFFICIENT OF VARIATION
LOT | RACE ; a 7 ‘ ian
Non- ; Abs. Rel. dif- Non- - Abs. Rel. dif-
conjugants| CoMugants difference | ference | conjugants | Conjueants difference ference
A. Wild cultures
| | | % I | | %
1 | 24.57+0. 62) 15.28+0.38 9.29+0.73) 37.8 11.03+0 28) 7.630. 19) 3.35+0.34 30.4
2 13.84£0.71| 13.20+0.37, 0.64+0.80) 4.6 8.43+0.43) 8.00+0.23) 0.434049) 5.1
3 | 24.20+0.87| 13.54+0.50 10.66+1.00 44.0 10.60+0.39) 6.91+0.26) 3.69+0.47 24.8
4 | 25.42+0.98 9.56+0.50 15.86+1.10 62.4 16.36+0.65) 6.86+0.36 9.5040 74 58.1
5 | 19.08+0.91) 7.90+0.94 11.18+1.3i; 59.0 i 13.27+0.64 6.05+0.72) 7.22+0.96 54.4
6 | 20 970.56 13.32+0.39 7.65+0.68 36.5 | 11.270 31) 7.34+0.21) 3.93+0.37) 34.9
7 17.43+0.73| 11.66+0.44 5.77+0 85, 33.1 9.57+0.40) 6.91+0.26) 2.65+0 48) 27.7
8 10.89+0.69) 7.02+0.46 $.87=0:83) 35.5 7.31+0.47, 5.20#0.34) 2.1140.58 28.9
B. Pure races
9 } ¢ 18.38+0.62| 11.13+0.34, 7.45+0.71) 40.6 11.58+0.40| 7.39+0.22) 4.19+0.46 36.2
10 | k 9.65+0 70) 7. 140.47) 2.51+0.84 25.9 6.89+0.50) 5.61+0.37) 2.28+0.62, 33.1
11 k 11.64+0.56) 8.87+0.80) 2.77+0.98) 23.8 || 8.70+0.42) 7.60+0.69) 2.10+0.81) 24.1
12 | k 14.25+0.68 7.51+0.29) 6.74+0.74 47.3 | 9.91+0.48) 6.05+0.23) 3.86+0.53, 39.0
13 k 14.35+0.65) 10.96+0.29) 3.39+0.74 23.7 | 10.23+0.49 8.46+0.22) 1.77+0.54) 17.3
14 | Nfz | 7.92+0.65) 5.53+0.41) 2.39+0.77 30.2 5.48+0.45) 4,030.30) 1.45+0.54! 26.5
15 | Nfe | 18.18+1.56) 6.66+0.45 11.52+1.62, 63.4 || 12.31+1.07) 5.01+0.34) 7.3041.12 59.3
16 Ca | 15.27+0.73) 11.97+0.33) 3.40+0.80) 22.1 11.45+0.55) 9.30+0.26 2.15+0.61 18.8
17 C2 | 13.53+0.62) 11.46+£0.47| 2.07+0.78) 15.3 10.23+0.47| 9.40+0.39) 0.83+0.88 8.1
18 | g 17.97+0.86 7.80+0.29 10.17+0.91) 56.6 13.68+0.67) 6.31+0.23) 7.37+0.71, 53.9
19 | 9 | 16.49+0.72) 10.11+0.37, 6.38+0.81) 38.7 12.18+0.54) §.55+0.31) 3.53+0.62) 29.0
C. mixed
20 | Cr4s) 24.71£0.97) 10.31+0.44| 14.4041.07) 58.3 | 20.1840 82, 7.95-+0.34 12.23+0.99| 60.6
21 La+k| 41.67+1.59| 6.68+0.32) 34.99+1.62) 84.0 24.07+0.97 5.56+0.27 18.511.01) 76.9
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, NO. 1
18 H. S. JENNINGS
a greater coefficient of variation. This suggestion contained
perhaps the germ of a correct explanation, but erred in empha-
sizing a supposed differentiation of the gametes from ordinary
adults. Pearl showed that in cultures which are not conjugating,
nor near a period of conjugation, the coefficients of variation are
as great as for the general population of those containing conju-
gants (’07, p. 231), and my own extensive data (’08) confirm this
fully. In such cultures there are no gametes and no specimens
in the process of differentiation into gametes, so that heterogene-
ity on this score cannot account for their greater variation as
compared with the conjugants.
2. Equalization
Pearl (’07, p. 262) discusses the possibility that a process of
equalization has occurred during conjugation; ‘“‘that the pro-
conjugants were simply a random sample from the general pop-
ulation having equal variability with it,” and that the decrease
in variability is due to ‘‘a pronounced tendency toward equali-
zation in size of the two members.’”’ He adduced evidence to in-
dicate that such a process of equalization does not occur to any
considerable degree. A tendency to equalization does exist, as
we shall see later, but (as will later appear) it is not of such a
character or degree as to account for the greatly decreased varia-
bility and smaller size of the conjugants. Moreover, as we shall
immediately see, such an explanation is gratuitous, since there is
a fully adequate explanation on other grounds.
3. Growth
In my paper of 1908, I showed that a large proportion of the
variation in an ordinary culture of Paramecium is due to growth.
A random sample of the population includes young individuals,
that are very small; individuals in all stages of growth up to the
largest sizes and individuals that have again decreased in length
preparatory to fission. Now, the conjugants include neither the
young, small individuals, nor the largest ones. Hence they show
CONJUGATION IN PARAMECIUM 19
much less variation than the population as a whole. The same
thing would be found true (in possibly a less degree) for man or
any higher animal; mated couples would be found on the whole
less variable than a general sample of the population that included
children. In one respect the condition in the infusorian is
peculiar; the conjugants do not grow so large as the individuals
that are to undergo fission without conjugation. Thus the conju-
gants represent a rather definite, limited stage of growth, exclud-
ing the extremes at both ends. That this is fully sufficient to
account for the lesser variability of the conjugants is demonstrated
by the fact that non-conjugants at a definite growth stage show as
little variability as do conjugants. The coefficients of variation
in length for conjugants range as a rule from 5 to 8 per cent, as
shown in tables 1 and 2, and in Pearl’s tables. In my paper of
1908, I showed that individuals beginning fission (and therefore
at a fairly definite growth stage) show coefficients of variation as
small as those for conjugants, and differing as much from those of
the general population (p. 454). Coefficients of variation as
low as 4.5 were found for samples of a definite age. The low vari-
ability of conjugants is then fully accounted for by the fact that
conjugation does not occur till a certain stage of growth has been
reached; and that conjugation occurs before the animals have
reached the largest size, that precedes fission.
One of the most striking things to be observed in a conjugat-
ing culture is the existence along with the conjugants of many
individuals of much greater size than the conjugants. This will
best be seen by comparing the measurements of conjugants and
non-conjugants of given lots, as exhibited in tables 34 and 35 of
the appendix; it is indicated in tables 1 and 2 by the fact that
the range of variation invariably extends to much higher limits
in the non-conjugants than in the conjugants, as well as by the
fact that the mean is higher for the former. Fig. 2 shows a col-
lection of conjugants and non-conjugants from a culture of the
race k; the specimen marked e shows one of the very large non-
conjugants. It is easy to isolate from a conjugating culture many
non-conjugants that are larger than any of the conjugants.
20 H. S. JENNINGS
Increase in size and variability of the conjugants before fission.
Further evidence as to the significance of the decreased size of
the conjugants will be reached by asking the following question:
Do the conjugating individuals remain smaller than the non-con-
jugants; or do they increase in size before they divide, till they
finally reach the size of the large non-conjugants? A parallel
question may be asked regarding the variability of the conju-
gants; does this increase to the normal amount before the ex-
conjugants divide?
A 7 Li
a b c d e
Fig. 2. Conjugants (b and d) and non-conjugants (a, c and e) from the race k,
showing the relative sizes. > 333.
On these points observations were made for a number of cul-
tures; those for lot 6 (of table 1) are the most complete. From
this lot I selected fifty of the largest non-conjugants, all larger
than any of the conjugants, and compared their dimensions with
the dimensions of conjugants about 86 hours after separation;
also with random samples of conjugants and non-conjugants.
The results are given in table 7.
As this table shows, the conjugants before dividing increased
in size till they were fully equal to the selected largest non-con-
CONJUGATION IN PARAMECIUM 21
jugant individuals. The mean dimensions of the conjugants
increased before fission from 181.49 by 48.11, to 214.48 by 64.24.
These later dimensions of course exceed greatly the mean for the
non-conjugant population.
More extensive data on the increase in size of the conjugants
before fission, though without comparison with the non-conju-
gants, is given in table 19, on page 69. Here the conjugants of
the two wild lots (22 and 23) had increased in length 20.62 per
cent and 11.76 per cent, respectively, after separation; those of
TABLE 7
Changes in size and variability of the conjugants before they divide, in comparison with
non-conjugants, from lot 6 (wild culture)
LENGTH BREADTH
DATE :
Standard
Standard. oem
deviation
Coefficients
Re; Mean deviation of
°
variation variation
Jan. 30, '08 Non-conju-
gants, ran-
dom sample. .) 318186.10+0.79 20.97+0.56 11.27+0.31) 200) 49.98+0.49
Feb. 1, '08 Largest non-
conjugants . £0211.52+1.99 20. 84+1.41) 9.85+0.67) 50 62.48+1.17
Jan. 30, '08Conjugants, | |
random sam-
io) ee Gnaeritad | 272151.49+0.54 13.32+0.39, 7.34+0.21) 38, 48.11+0,57
Feb. 1, ‘08Conjugants, 36
hours after
separation....| 50.214.48+2.09 21.92+1.48 10.22+0.70| 50 64.24+1.06
So
21+0.34, 20.43+0.72
wo
290.82, 19.66+1.37
94+0 61 16.51+1.31
=
09+0.75) 17.27+1.20
race k (lot 24) had increased 21.82 per cent. Comparing the
size of the separated conjugants of race k as given in table 19 with
the size of the non-conjugants as given in table 2 (lots 10-13),
we find that the corjugants have become before fission fully
as large as the non-conjugant population.
Tables 7 and 19 show also that the variability of the conjugants
likewise increases considerably after they separate. In lot 6
(table 7) the variability in length of the separated conjugaats
approaches that of the non-conjugants. The coefficients of
variability for the separated conjugants of table 19, while con-
siderably larger than those for the united conjugants, are some-
what below those usual for the non-conjugating population, as
22, H. S. JENNINGS
shown in tables 1 and 2. There is of course an intelligible reason
for their variability remaining less than that of the non-conjugant
population, for the latter includes young individuals as well as
old ones, while the ex-conjugants consist entirely of old individ-
uals. There is no indication of a real differentiation of the con-
jugants from the non-conjugant population in respect to varia-
bility save as a result of this exclusion of young specimens from
the conjugants.
The general conclusion would therefore seem to be Justified
that the conjugants are not differentiated from the non-conjugants
of the same race, save temporarily, the differences disappearing
practically before the first fission of the conjugants.
4. Racial differences
We have seen the cause of the lessened variability of the con-
jugants within a pure race. Another ground for less variability
in the conjugants, when the culture is not limited to a single
race, is seen on examination of the data for the mixed cultures
(lots 20 and 21, table 2). Here the coefficients of variation for
the conjugants are respectively 60.6 per cent and 76.9 per cent
less than that for the non-conjugants; or to put it another way, in
lot 20 the variability of the non-conjugants is 2.54 times as great
as that for the conjugants; in lot 21 it is 4.33 times as great. The
remarkable differences become evident to the eye on comparing
the measurements of the conjugants and non-conjugants of these
lots, as shown in tables 34 and 35 of the appendix. These great
differences in variability are due to the fact that but one of the
two races present conjugated; so that the non-conjugant sample
included members of two diverse races, the conjugant sample
but one. As shown elsewhere (Jennings ’10), it frequently hap-
pens that in a mixture of races but one race conjugates. This must
often greatly affect the relative coefficients of variation in wild
cultures of unknown racial composition.
We conclude therefore that the less variability of conjugants
as compared with non-conjugants is due (1) to the fact that the
conjugants include only a limited number of growth stages, inter-
< CONJUGATION IN PARAMECIUM 2B
mediate between the largest and the smallest; (2) to the fact that
in mixed cultures not all the races conjugate at the same time.
The consequences of the lessened variability of the conjugants
will be considered later.
ASSORTATIVE MATING
Pearl’s explanation
Pearl (07) concluded from his study of conjugating Paramecia
that there is a marked degree of assortative mating in these
animals, z.e., that large individuals tend to mate with large
ones, small individuals with small ones. This is an extremely
important matter (as Pearl well recognized) for the understanding
of heredity, variation and evolution in these organisms, and we
must examine into the matter with care. Is there actually assort-
ative mating? What are its degrees and limitations; what are
its causes; what its effects in heredity and variation?
In the study of these matters the method used by Pearl, and
the one we shall to a large degree follow him in using, is to study
the correlation between the members of pairs. We shall however
endeavor to put to the test of experiment such questions as are
open to it, and to combine the statistical and experimental study
with the results of direct observation. We shall take up first
Pearl’s explanation of how assortative mating occurs; then give
an account of direct observations of the process of conjugating,
in their bearing on this explanation. ‘ We shall then enter upon
a statistical and experimental analysis of the facts.
Pearl showed that when we measure the individuals making
up the pairs in a Jot of conjugants, there is a rather high correla-
tion between the two members; that is, large individuals are found
mated with large, small with small. The details and degrees
of this we shall take up later. Pearl’s explanation of this cor-
relation is that there is a real assortative mating—larger indi-
viduals mating with larger, smaller with smaller. The way in
which this assortative mating is brought about, according to
Pearl, is essentially the following: In a typical conjugation the
24 H. S. JENNINGS
two individuals first place their anterior ends together; these
adhere. Then the two bodies are brought in line, and if the two
mouths come in contact, they adhere, and conjugation becomes
complete. Now it is evident that if the two animals are not of
approximately the same length, mouth and anterior tip will not
both come into contact, and conjugation will therefore not be
completed; in this case ‘the individuals separate again or die, and
no conjugation results.” Hence it is only individuals of approx-
imately the same size that will conjugate.
This explanation of Pearl’s is based to a certain extent on
observation of the process of conjugation, andits essential correct-
ness seems highly probable. But Pearl himself was able to make
but few observations on the behavior in the process of conjugat-
ing (as he notes on p. 266 of his paper), and it will be well there-
fore to add what we can along this lme. Some of the details
are of much importance for understanding the limitations as well
as the potentialities of the assortative mating. We shall examine
first the typical cases, then some of the variations.
OBSERVATION OF THE PROCESS OF CONJUGATING
The first contact between the individuals about to conjugate
is very commonly at the anterior tips, and I am able to give figures
of a number of pairs in which the process had gone no farther than
this (fig. 3, a, b,c). The two tips where they meet form projec-
tions and depressions, which interdigitate and hold the two
together. This I have often noticed in the living specimens, and it
is indicated in fig. 3, at b. Then the bodies themselves are brought
together. There is at first no union except at the anterior tip,
until the mouths are reached. These then unite, so that the ani-
mals are held together at two points only. This stage in the
process is represented by e, f, g (fig. 3), drawn from preserved
material.
Then the bodies become closely applied to each other through-
out the stretch from anterior tips to mouth, and for a certain dis-
tance behind the mouths. Apparently however the union is
not so firm elsewhere as at the anterior tips and the mouth;
CONJUGATION IN PARAMECIUM 25
A
DD
Dy
Fig. 3 Characteristic attitudes in conjugation. All are drawn with the pro-
jection apparatus from fixed specimens, save h, which is a free-hand sketch from
a living pair. a, e, f belong to the race k; c, d, g, i, 7 to the race C2; b, k and lL
to the race c (all aurelia); a, b, and c, stages with only the anterior tips in contact;
e, f, g, anterior tips and mouths in contact; h, reversed pair, at beginning of con-
jugation; 7, 7, most complete union; k, view of pair in profile, showing the crossing
(also seen in 7); 1, last stage, just before complete separation. Magnification (for
all but h), 333 diameters.
26 H. S. JENNINGS
specimens not in contact elsewhere than at these points are fre-
quently seen. These facts are noted by Pear] (’07).
But this state of full contact is by no means reached without
variation and active movements. On the contrary, there is
frequently a period of twisting, turning, contracting, and shift-
ing about, before the final result is reached. Certain points are
important.
1. At first apparently only the anterior tip (and possibly the
mouth) is adhesive. The anterior tip of one specimen may ad-
here, at least temporarily, to almost any part of the body of an-
other; certainly to any part of the oral surface. Thus one often
sees most irregular adhesions, the first specimen perhaps trans-
verse or oblique to the second, adhering by its tip to the middle
of the oral groove of the second; or one specimen is attached to
the posterior half of another, trailing behind it; or three or four
may be irregularly attached. Sometimes the animals become
attached in reversed position—the anterior tip of one to the pos-
terior part of the other, as in fig.3,h. Such irregular attachments
have frequently been described and figured.
2. If irregularly attached, the animals begin to pull, twist,
contract, and shift, till the relative positions are many times
changed. Not rarely one sees a pair completely separate, to
reunite in a new position. More often they remain in contact,
but the relative position and the parts in contact are changed
by gliding and pulling. I have frequently seen pairs in the re-
versed position of h, fig. 3, that finally came into the normal posi-
tions, and underwent a typical conjugation. On the whole the
period of ‘fitting’ in conjugation is one of great and varied activity.
3. Among the movements at this period of ‘fitting’ are con-
tractions and bending. One gets the impression that the animals
are making an active ‘effort’ to bring the mouths into contact.
A certain amount of curving of the anterior tips is common even
before the mouths have come in contact. If when the bodies
are brought in contact the mouths do not match, the curving and
bending becomes very marked as in e, f, g, h, fig. 3. This is of
course most necessary when the two individuals are not of the same
length; the longer then may become curved (as in f and g, fig. 3,
a
CONJUGATION IN PARAMECIUM 27
and in e and fh, fig. 15); so that a considerable degree of equaliza-
tion in length may occur. In measuring the avimals it is difficult
to detect or allow for this curving, as the animals in conjugation
are slightly oblique to each other in any case (owing to the oblique-
ness of the oral grooves). In actual practice, most such equalized
pairs are measured by simply taking the distance from anterior
to posterior tips, in the two specimens. This tends of course
to produce a correlation that did not exist before the union.
It is however important not to exaggerate the generality or
amount of this equalization, and especially to remember that it
is only one phase of a process of fitting, the remainder of which
would lead to real assortative mating.
4. All of this shifting and contraction may be insufficient for
producing a proper fit: in such cases the animals separate. Such
separation is often observed. This is of course an essential point
for producing real assortative matiog. It appears clear that
individuals of nearly the same size musi fit readily, and that the
more unequal they are, the less likely they are finally to fit and
remain united.
5. From the description thus far, it is clear that the first attach-
ment may not be at the anterior tips of both individuals. It is
more likely to be here, because both anterior tips are adhesive,
while most of the rest of the body is not. But the anterior tip
of one may come in contact with the oral surface of the other
some distance behind the anterior tip of the latter. If the two
animals are equal, of course the two mouths will now not come
together, until the positions have been shifted; but if the two ani-
mals are unequal,—if the one lying more to the rear is shorter,—
the mouths may then come in contact, and complete conjugation
will take place between unequal specimens. Thus the assorta-
tive mating, or the correlation in size between two members of a
pair, cannot be expected to be complete, since many unequal
pairs are found. A number of such are shown in fig. 15, page 53.
Thus on the whole direct observation of the process of conju-
gation and of the conjugated pairs is favorable to Pearl’s view
that real assortative mating occurs, and to his explanation of the
way it occurs. One important point, not brought out by Pearl,
28 H. S. JENNINGS
results from our description,—namely, that there is a real tend-
ency toward equalization in length of the two members of the
pair. This of course tends to produce correlation where it would
otherwise not exist; in other words it gives, so far as it goes, an-
other explanation of at least a portion of the correlation, and one
which would deprive the process of its significance for variation,
heredity and evolution. So far as direct observation goes, how-
ever, this factor may be of very slight value, accounting for but
little of the observed correlation, but it must be kept in mind in
the statistical and experimental work, and evidence as to its real
value obtained if possible.
CORRELATION IN SIZE IN THE MEMBERS OF PAIRS OF CONJUGANTS
We will now examine the facts as to the correlation in size of
the members of a number of lots of conjugants. The data are
given in tables 1, 2 and 3, pages 9 to 11, table 3 giving the facts for
a number of lots in which only conjugants were measured, while
tables 1 and 2 relate to lots in which a comparison was made be-
tween conjugants and non-conjugants. In all these tables the
data are for conjugating pairs measured while still united; later
will be found the facts regarding correlation in pairs measured
after separation (tables 19 and 20).
MEANING OF THE COEFFICIENTS OF CORRELATION
In all these tables the degree of correlation is stated in terms of
the coefficient of correlation; coefficients will be found varying
from —0.193 to + 0.940. In order to grasp clearly what is
meant by correlation, and by the different coefficients of correla-
tion, it will be helpful to examine the pairs shown in fig. 4, and
the diagram of fig. 5.
Fig. 4 shows a number of pairs selected in such a way as to
exhibit the condition of affairs in the case of marked positive cor-
relation. Where one member of a pair is large, the other is like-
wise large, and at the other end of the series both members are
PARAMECIUM
IN
ATION
x
I
CONJU
‘sdajauIVIp GZ ‘UOTYVOYTUsSe IY
‘yoyered ose ,g-g pue Q-V seat oy “seyeUT ][VUIS sO [[VUIS ay} ‘S97 aZIv] SUIAVY S[ENPIAIPUI oFAe] OY} “OZ1S UL WOTPBTOLIO9
aAT}ISOd ay} 4IGIYxXe 0} SB posed os “go6T ‘8 Youve ‘(eljaine) %) 9ov1 oyz Jo eINy[Nd B WLOTT sired jo dnois poyopes YF “B14
LO
va
30 H. S. JENNINGS
small. Thus if we select the pairs according to the size of one
member only, we shall nevertheless find the other member to
have nearly the same size.
If in all the pairs of varying size the two members of each pair
were precisely equal in length, then the correlation would be com-
plete; the coefficient of correlation would be unity (1.000).4 But
what is meant by coefficients of correlation less than 1.000; by
such coefficients as 0.398 (lot 1, table 1), or 0.507 (lot 3, table 1)?
And what is the difference between the two latter cases? Why
should lot 3 be given precisely the coefficient 0.507, in place say
of 0.275 or 0.860? It is this question that the diagram of fig. 5
is intended to assist in answering.
Suppose we take all the 360 individuals forming the 180 pairs
of lot 1 (table 11, page 47) and group them according to their
lengths into groups at intervals of four microns. We thus ob-
tain 29 groups, the smallest 148 microns long, the largest 260
microns long. We then place these 29 groups side by side, be-
ginning with the largest and proceeding to the smallest, as in fig.
4. But instead of giving the actual outlines as we did in fig. 4,
we use merely lines giving the length of each group. This gives
us the vertical lines of fig. 5,—the length of the largest individuals
being represented by the line A-B, that of the smallest by the line
C-D, the others by the intermediate lines. Then the ends of these
lengths form the oblique line B-D. The average length of
all is shown by the line 0-0’.
Now suppose we examine the mates of the individuals of these
groups—getting the average length of the mates for each group.
If correlation were complete (coefficient 1.000) the mates would
be of the same lengths as the first individuals (which we may call
the principals) ; their upper ends would lie on the same line B-D.
4It is worthy of special notice that in the particular case of correlation with
which we are dealing, where the two members are of the same sort, this absolute
equality of the two members of the varying pairs is the necessary condition for
producing the coefficient 1.00. In other cases of correlation this is usually not the
case. Thus it would be possible to study the correlation between the bodily stat-
ure and the length of the little finger in man. In this case complete correlation
(coefficient 1.00) would mean that the same proportion of one to the other was
present in all the varying cases.
260
240
220.
180
160
140
120
100
80
60
40
20
CONJUGATION IN PARAMECIUM 31
A Cc
Fig.5 Diagram to illustrate correlation in size between the members of pairs in
lot 1 (see explanation in the text and compare fig. 4). The vertical lines represent
the lengths of individuals of the different classes arranged in order from longest
(260 microns) to shortest (148 microns), their terminations tracing the oblique
line B-D. The distance from A-C to O-O’ shows the average length for all. The
distance from A-C to the broken line H-F shows the lengths of the mates of the
classes of individuals represented by the vertical lines. The general trend of this
line E-F is shown by the line X-Y (‘regression line’). The value of the coefficient
of correlation is given by the proportion which the line O-X is of the line O-B.
But in fact we find that the mates are not of exactly the same
lengths as the principals. Thus in lot 1 (table 11, page 47),
the two individuals having the length 160 microns (40 units)
32 H. S. JENNINGS
have mates respectively 164 microns and 196 microns in length,
so that the average of the mates is 180 microns. We thus work
out the average length of the mates for each length of the prin-
cipals; this gives us the results shown in table 8. As this table
shows, the length of the mates increases on the whole as the
length of the principals increases (though not at the same rate),
so that larger individuals have somewhat larger mates.
Now we mark on the diagram of fig. 5 the lengths of the mates
corresponding to the lengths of each group of principals repre-
sented by the vertical lines; we find that these lengths of the mates
TABLE 8
Mean lengths in microns of the mates for the individuals of diverse given lengths, in
the 180 pairs of lot 1
NUMBER OF PAIRS LENGTH OF PRINCIPAL MEAN LENGTH OF MATES
1 148 188.0
1 152 192.0
2 160 180.0
4 164 181.0
2 168 182.0
4 172 192.0
11 176 188.4
23 180 192.3
19 184 190.0
29 188 194.9
36 192 194.2
35 196 198.9
41 200 200.6
38 204 202.4
34 208 204.2
31 212 205.0
15 216 207.5
15 220 201.6
6 224 205.3
4 228 206.0
4 232 202.0
1 236 248.0
1 240 228.0
2 248 222.0
1 260 212.0
A CONJUGATION IN PARAMECIUM 33
are distributed on the irregular (heavy) line H-F. The course
of this line shows that the smaller individuals, near the side
C-D, have mates larger than themselves; that the larger individ-
uals (near A-B) have mates smaller than themselves, while the
intermediate individuals have mates of nearly their own size.
But it is on the whole clear that the line H-/' does slope a little
in the same direction as B-D, only less; large individuals in the
left half of the diagram do have larger mates than the smaller
individuals, in the right half. That is, there is a certain degree
of positive correlation between the size of individuals and the size
of their mates. To show how marked this is, we may draw a
straight line X-Y, showing the general trend of the slope of the
broken line E-F. (The method by which this line is drawn will
be taken up later.) The line X-Y shows approximately what
would be the course of the line H-F if we had an infinite number
of cases; the irregularities in H-F’ are due to the limited number
of pairs with which we must deal. We may therefore look upon
X-Y as showing us the real mean lengths of the mates of the
individuals having the lengths shown by the vertical lines.
The position of this line X-Y with relation to the position of the
line B-D is now the important point for determining the degree
of correlation. We see that X-Y rises above the mean (O-O0')
where B-D rises above it, and falls below the mean where b-D
falls below it, thus sloping in the same general direction as 5-D.
If X-Y did not slope with B-D, but were instead quite horizon-
tal (coinciding with O-O’), then there would be no correlation
(coeflicient 0), since this would show that small individuals and
large individuals had mates of the same average size. On the
other hand, if X-Y not only sloped in the same direction as B-D,
but actually coincided with it (so that all specimens had mates
equaling them in size), then the correlation would be complete
and the coefficient would be 1.00. But as a matter of fact X-Y
falls neither at O-O’, nor at B-D, but between the two,—and
its precise position is what determines the numerical value of the
coefficient of correlation. The line X-Y cuts off at X just 0.398
of the entire distance from O to B (that is, it cuts off 0.398 of the
angle between the lines O-O’ and B-D); therefore the coefficient
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. Il, NO. 1
34 H. S. JENNINGS
of correlation is 0.398. If X fell half way between O and B, the
coefficient of correlation would be 0.500; if it fell nine-tenths of the
distance from O to B, the correlation would be 0.900, ete.°
In place of measuring the proportion of the distance O-B cut
off by X, we could of course measure on any of the vertical lines
of the diagram the portion of the distance from the line O-O'
to the line B-D that is cut off by X-Y; the result would be the
same.
Fig. 5 may be used further to illustrate negative correlation.
If the line X-Y sloped in the opposite direction from B-D, falling
below O-O’ where B-D rises above it, this would of course show
that the larger the individual the smaller its mate; 7.e., we should
have negative correlation. To produce complete negative cor-
relation (coefficient, —1.00) the line X-Y would make the same
angle below O-O’ that B-D makes above it, and vice versa. The
degrees of negative correlation would then be determined in the
same way as those of positive correlation.
All this may be clearly illustrated if we make a diagram rep-
resenting only the upper part of fig. 5, above D (that is, including
only. the varying portions of the lines); such a diagram is given
in fig. 6. On this diagram are shown the various positions of
the line X-Y corresponding to different degrees of positive and
negative correlation when the heavy line B-D shows the dimen-
sions for the principals (also complete positive correlation in the
mates).* (The diagram in fig. 6 is made on a somewhat dilfer-
ent scale from that of fig. 5, the vertical distances being greater
5 This exposition would not hold, in its present form, for eases where we seek
the correlation between unlike things (as between the stature and the length of
the finger in man). In such cases the coefficient of correlation depends partly on
the relative variability of the two sets of things compared. In the case of corre-
lation between likes, with which we are dealing (where in fact the two classes
compared are composed of the same individuals), this complication does not come
in; the means, standard deviations, and coefficients of variation are the same for
the two classes, so that the coefficient of correlation is identical with the coefficient
of regression. For such cases our exposition holds without modification.
® Some authors report coefficients of correlation greater than 1.00. This is of
course due to arithmetical errors, since when the measurements all fall on the same
straight diagonal line passing through the mean, the coefficient is but 1.00, and it
is decreased when any of the measurements fall elsewhere than on this line.
CONJUGATION IN PARAMECIUM 39
i.
.9
8
7
NG
Es
a \
3 NN
SSA
a2 SSN
RB BSS S
2 ===
sesS5
te Liz
| Gi
ay
as 7)
~.6
237
-.9
=a
Fig. 6 Diagram illustrating the significance of different coefficients of correla-
tion (should be compared with fig.5,tothe upper portion of whichit corresponds).
The vertical lines are the terminal portions of the lines representing the lengths of
various classes of individuals, arranged according to their size, the largest (A-B)
at the left. The terminations of these diverse lengths trace the oblique line B-D.
O-O marks the position of the mean length of all. The lines B-D and O-O and the
other oblique lines show the positions of the theoretical average dimensions for
the mates of individuals of the given sizes, in case of different coefficients of cor-
relation. If the dimensions of the mates fall in the same line B-D as those of the
principals, the correlation is 1.000; if they fall at the mean O-O the correlation is
0; if they fall at right angles to B-D, the correlation is —1.00. If they fall in a
line dividing O-B into equal parts, the correlation is 0.500. The coefficients, from
—1. to +1., by tenths, are illustrated in the diagram, the numbers at the
extremity of a given line showing the coefficient to which it corresponds.
in proportion to the horizontal distances.) We shall use similar
diagrams for showing the correlation in the various lots studied.
From figs. 5 and 6, it will readily be conceived what is meant
when such coefficients of correlation are mentioned as appear in
the last columns of tables 1, 2 and 3.
The position of the line X-Y is determined in practice simply
by finding the coefficient of correlation, then marking off the
equivalent proportion of O-B above O, and of O’-D below 0’,
36 H. S. JENNINGS
and connecting these two points by a line. (In the case of cor-
relation between unlike things, modification of this procedure
would be necessary; a coefficient of regression is derived from the
coefficient of correlation, and this gives the position of the line
X-Y. But in such eases as we are dealing with the value of the
coefficients of correlation and of regression are the same.)
The method of finding the coefficient of correlation is of course
described in text-books of statistical methods. For computing
correlation under the particular conditions with which we are
here dealing, an improvement over the usual methods is described
in a recent paper by the present author (711).
TABLE 9
Coefficients of correlation in length between the members (A and B) of pairs in the
different classes of cultures of Paramecium (compare tables 1-3)
WILD CULTURES, OF MIXED RACIAL E
JRE RACE:
COMPOSITION PU CES
ae 2 ge.
eee 2 2 2s e23
1 11 | 180 0.398+0.030 9 c 43 125 0.132+0.042
2 36-142 0.268+0.057 | 10 k Ad 26 —0.193+0.090
3 37 82 0.507+0.049 11 k | 45 14. —0.137+0.125
4 38 42 0.499+0.055 | 12 k 46 78 0.367+0.047
6 39 | 136 0.428+0.033 | 18 k 47 =: 168 0.184+0.036
7 40 7 0.333+0.048 || 24 k 122 0.210+0.041
22 | 55 | 102 | 0.359=0.041 | 14 Nfe 21 0.295+0.095
23 | 22 | 148 | 0.245+0.037] 16 Nps 25 0.257+0.089
and 16 G 48 150 0.507+0.029
23 17 Cs 49 69 | 0.318+0.052
18 g 50 84 0.251+0.049
Mixtures of two species, both con- 19 g 51 87 0.323+0.046
jugating
26 31 | 0.939+0.010 | |
27 10-170 0.940+0.004 |
Mixtures of two species, only one
conjugating
21 54 49 0.408+0.064
lod
CONJUGATION IN PARAMECIUM 37
CORRELATION IN DIFFERENT CLASSES OF CULTURES
Turning now to an examination of the coefficients of correla-
tion of the various lots, as shown in tables 1, 2 and 3, we find that
we may distinguish four different classes: (1) wild cultures,
of unknown racial composition; (2) wild cultures known to con-
tain pairs of the two species, aurelia and caudatum; (3) cultures
consisting of a single pure race; (4) mixtures of two known races.
It will be well to consider these separately. The coefficients of
correlation for these classes are summarized in table 9.
Wild cultures of unknown racial composition
Tables 1, 3 and 9 give us the correlation for eight lots of this
class. We find positive correlation in every case, the coefficients
ranging from 0.245 to 0.507, with an average for the entire eight
B
D
Fig.7 Diagram illustrating the extent of correlation in the eight ‘wild’ cultures
of tables 1 and 3 (at a to 6), and in the mixture of two species in lot 27 (atc). Be-
ginning at a the lines show in order the coefficients of correlation for lots 23, 2, 7,
22, 1, 5, 4, 3, (at b); then that for the mixture of two species at c.
O-O, mean and line of no correlation; B-D, line of complete positive correlation
(compare figs. 5 and 6).
38 H. S. JENNINGS
of 0.380. In the five series studied by Pearl the coefficients were
higher, ranging from 0.430 to 0.794, with an average of 0.614.
Fig. 7, a to b, illustrates the correlation in our eight wild cultures.
The actual relations of the individuals of various lengths to their
mates are shown for lot 1 in table 11 and fig. 5; for lot 3 in table
37 and fig. 8.
B
1 240
230
220
ail 160
Fig. 8 Diagram illustrating correlation in lot 3 (‘wild’ culture). B-D, lengths
of the classes of individuals in order of size. H-F, lengths of the mates for these
classes. X-Y, regression line showing the general trend of the broken line E-F,
and indicating by the proportion of the line O-B which it cuts off, the value of the
coefficient of correlation (0.507). O-O, mean length for the entire lot. The num-
bers at the right show lengths in microns.
Since we now know that wild cultures often contain a number
of races of diverse sizes, the question arises whether the corre-
lation may not be a result of this fact. If members of large races
conjugated only with other members of large races, and members
of small races likewise mated only together, there would result
a positive correlation, provided that more than one race were
undergoing conjugation at the same time. If such interracial
selectiveness were the only basis for the correlation, we should
CONJUGATION IN PARAMECIUM 39
of course find no correlation on studying conjugating pairs that
all belonged to a single pure race. To get light on this matter,
we may examine the correlation (a) in cases where there were
known to be two greatly differing races (two so-called species) ;
(b) in cases where members of but one race are present.
Cultures containing pairs belonging to two different species
In two cases I was able to obtain conjugants from cultures
containing both Paramecium caudatum and P. aurelia. These
two species, as is well known, differ considerably in size, but very
little in other external features.’ What happens when members
of the two species, mingled together, conjugate at the same time?
Simultaneous conjugation of the two was obtained as follows:
Material known to contain Paramecium caudatum in large num-
bers was brought into the laboratory, and mixed with cultures
of the aurelia race k. About a week after the mixture was made,
the conditions became favorable for conjugation, and both species
mated. The resulting matings are shown, for the culture from
which the largest numbers were measured, in table 10.
From this table it is clear that the large individuals of caudatum
mated exclusively with other caudatum; the small individuals
of aurelia only with other aurelia. The two cultures of this sort
that were examined gave coefficients of correlation in length,
of 0.939 +0.010 and 0.940 = 0.004, respectively ; so that the cor-
relation was almost perfect. Careful examination of all the
pairs measured gave no single case in which it appeared, to the
practiced eye, that caudatum had mated with aurelia.
Thus when caudatum and aurelia are present together in a cul-
ture, they do not intermix in conjugation,—certainly not to any
marked extent, and apparently not at all. (Simpson, ’01, saw
two cases of what he believed to be conjugation between aurelia
and caudatum. The ex-conjugants died after one fission. It is
of course possible that crosses might be induced by proper isola-
7 For a detailed account of the differences between the two, see Jennings and
Hargitt, 710.
40 H. S. JENNINGS
tion, even though they do not occur in nature, or occur there but
rarely).
Our results with two species are then favorable to the idea that
correlation is produced by members of related races conjugating
together. We now turn to the records for conjugation within a
single race.
TABLE 10
Correlation table for the lengths of 170 pairs from a culture contain-
ing pairs of both aurelia and caudatum (correlation, 0.940 +0.004).
The unit of measurement is four microns (0.004 mm.,) (so that
the first pairs to the left have members measuring 112 and 108
microns in length)
eile | heal l Ree ae Sse
128.2930 31/32/33 34 3530137 3830-40 4142.43 44 45,46 47 48 49)50)51 52 53,54 55 56 57 58.59)
j i St Pe | | J | |
— tt SS Ses Se tes i —
| Beeea at 4
| || | Peers Pad lid
a de bo
el
a
DAR OWN NL & Ome
ae ee
ie)
ee)
=
— ee
_
i
ee
es Ce os
bots
Nww ree
Ney ke ee
ym bo to
~
-
Oe
_
| | |
| | | | | | |
55 | Ved yea | [1
|
35 510 710124M\10, 56 5 2.21 1 3l 1) 1) 9| ss! gal e| 2| 3| 5 1 1) 170
| ' 1
CONJUGATION IN PARAMECIUM 41
Conjugation within pure races
The coefficients of correlation in length for twelve conjugations
within pure races (all the conjugating individuals being derived
originally, in each case, from a single specimen) are given in tables
2, 3 and 9; they are illustrated in the diagram of fig. 9.
Examination of the tables and the diagram shows that
the correlation is indeed considerably less in the pure
1.8
-1.
A D
Fig.9 Diagram illustrating the correlation in the eleven lots composed of pure
races, of table 2. O-O, mean and line of no correlation; B-D, line of complete
positive correlation. a, lot 9; b, lot 10; c, lot 11; d, lot 12; e, lot 13° g, lots 15 and
18;h, lot 16;7, lots 17 and 19. Just beneath 7 is the line for lot 14.
caces than in the wild cultures. This is seen in. comparing
fig. 9 with fig. 7. From the first three coefficients of table 2
(0.182, —0.193, and —0.137) we might conclude indeed that sig-
nificant correlation is quite absent within a pure race. But on
examining the entire twelve lots from pure races, we find that this
conclusion will not hold. In all save the two lots just
mentioned the correlation is positive; and these two in which it
42 H. S. JENNINGS
was found negative are small, containing respectively but 26 and
14 pairs; they have been included in our account merely to illus-
trate the results reached if insufficient numbers are employed.
Other, much larger lots of this same race k (lots 12, 13 and 24 of
tables 2 and 3) gave significant positive correlation. Further,
we have in lot 16, from the pure race C;, a large lot (150 pairs)
giving the high correlation of 0.507 + 0.029.
We must conclude then that there is really positive correlation
in length in the members of pairs even when all belong to the
same race. The fact that it is much less in amount than in mixed
cultures is however of much significance. Fig. 10 illustrates
the very slight correlation in the large sample of lot 9 (race c)
while fig. 11 shows a similar diagram for lot 17 (race C,). The
average coefficient for the twelve lots from pure races is but 0.251,
as against 0.380 for our eight wild cultures, and 0.614 for Pearl’s
wild cultures. The difference in these averages is illustrated
in fig. 12.
The smaller correlation found in pure races, as compared with
mixed cultures, casts much light on the causes of the correlation.
It indicates most directly, of course, that in wild cultures indi-
viduals belonging to the same race, or to races of similar size,
tend to mate together. The bearing of this most important con-
clusion on other problems we shall bring out later. Here we shall
take up certain other evidence indicating this tendency of mem-
bers of the same or like races to conjugate together.
In wild cultures it is frequently found that members of diverse
races are conjugating at the same time. This is demonstrated
by isolating pairs of different sizes, and finding that they produce
progeny of permanently different characteristic sizes. We may
cite here a single case, described in an earlier paper (Jennings,
‘08, p. 494). Six races of diverse size (including the large race L,
with average length of about 200 microns and the small race C:,
with average length of about 125 microns) were derived from six
differing pairs of conjugants taken from a single culture January
29, 1908.
CONJUGATION IN PARAMECIUM 43
ae 200
160
F
x
a O
Né
140
-! 120
D
10
1
Fig. 10 Diagram for correlation (0.132) in the pure racec of lot 9. For expla-
nation, compare fig. 8, page 38.
Fig. 11 Diagram for correlation (0.318) in the pure race Cy of lot 17. Lettering
as in fig. 8, page 38.
44 H. S. JENNINGS
A D
Fig 12 Diagram showing the average correlation in length between the mem-
bers of pairs in wild cultures of one and of two species, and in pure races. O-O,
mean; also line of no correlation; B-D, line of complete positive correlation. E,
average correlation (0.250) for the twelve pure races of tables 2 and 3. F, aver-
age correlation (0.380) for the eight wild cultures of the present paper (tables 1 and
3). G, average correlation (0.614) in the five wild cultures studied by Pearl (’07).
H, average correlation (0.940) for the two lots of table 3 where two species were
present, both conjugating.
Mixtures of two known races
When cultures are formed by mixing two or more diverse
races of known characteristics, it is extremely difficult to induce
the members of both races to conjugate at the same time. I
kept many such mixtures for months but only once did I succeed
in getting both races to conjugate at once. This was in a culture
containing the races 7 and / (both aurelia). The race 7 is very
small, averaging but 95 to 100 microns in length, while x is larger,
averaging about 125 microns in length.’ The two races had been
living together in the same culture about five months (from
5 For measurements of these races, see Jennings ’08.
CONJUGATION IN PARAMECIUM 45
November 8, 1908) when conjugation was observed in the cul-
ture March 1, 1909. Unfortunately the conjugation was scanty,
so that only five pairs were found. Three of these measured
respectively 78 x 76, 78 x 76, 82 x 76 microns; the other two 138
x 138 and 138 x 140 microns (compare fig. 13). It is clear that
the first three pairs belonged to 7, the last two to k. Had we a
large number of pairs of these two races, it is clear that we could
form a correlation table from the culture as a whole that would
i
ip re Ay
k k k i i
Fig. 13 Conjugants and non-conjugants from a culture composed of a mixture
of the two aurelia races k and 7, showing that each pair is composed of individuals
of but one race,—either of k or 7. X 333.
show a high degree of correlation. The 7’s would all fall in the
upper left-hand corner of the correlation table, the k’s in the lower
right-hand corner.
Thus in this case the individuals of each race conjugated only
with members of their own race. In most cases where conju-
gation occurs in a culture containing two known races, it is lim-
ited to the members of one race, as I have already set forth
(p. 22). This fact that two races conjugate at different times has
of course the same tendency as assortative mating, in preventing
admixture of races.
46 H. S. JENNINGS
Before proceeding to further discussion of the causes and effects
of the correlation of the conjugating individuals, it is important
to examine certain other points.
TO WHAT IS DUE THE INCOMPLETENESS OF CORRELATION?
As we have seen (tables 1, 2, 3, ete.), correlation between the
members of pairs is never complete, but in most cases is a cer-
tain positive amount, less than 1.000. It is important to realize
as nearly as possible to what biological relations this corresponds.
Why, if there is correlation at all, should it not be complete cor-
relation?
Incompleteness of correlation may be due to a number of differ-
ent conditions, the more important of which we must consider.
1. Slight differences of less effect than great ones
Perhaps the most probable cause for incomplete correlation,
at least in such material as we are dealing with, would be this:
that slight differences in size between the members of pairs do
not prove a marked obstacle to their union, while greater dif-
ferences prevent their uniting. If this is the state of affairs,
we should find but slight correlation in collections or parts of col-
lections in which there are but small differences among the
individuals; higher correlations where the differences between
individuals are great. The highest degree of correlation would
be found in the case of a collection composed of two sets, those
of each set not differmg much among themselves, but the two
sets differing considerably.
That the condition we have sketched is the real condition in
the conjugation of Paramecium is shown in two different ways:
a. We have already seen that when the conjugating culture
contains two sets differing greatly (as when aurelia and caudatum
are both present), the correlation is very high; when numbers
of less differing races are present, as in ordinary wild cultures,
the correlation is less, but still marked; when only the closely
CONJUGATION IN PARAMECIUM 47
similar members of the same race are present, the correlation is
still less.
b. It is shown further when we compute the correlation for
portions of the various lots. Take for example lot 1 (tables 8
and 11.) Here we have 360 individuals varying from 148 to 260
microns in length; in the tables these are arranged in 29 classes
in order of size. The mean lies at almost exactly 200 microns.
TABLE 11
Correlation table for the lengths of the pairs in lot 1, with lines to show the ‘medium’
and ‘extreme’ pairs described in the text and illustrated in fig. 14.
(1) The pairs containing exclusively medium specimens are those in the small square
enclosed by the lines a-b-c-d (correlation 0.104). (2). Pairs at least one member
of which is of medium size (correlation 0.229) are between the lines c-g and c-e on
the one side; b-h and b-f on the other. (3) Pairs at least one member of which is
extreme (correlation 0.439) include all those outside the square a-b-c-d. (4) Pairs
consisting exclusively of extreme specimens (correlation 0.678) lie outside all the
lines drawn within the table.
The unit of measurement is four microns—so that the length 37 for example signifies
148 microns.
10014 6 9161319 21222312146 3400 1200 1/180
48 H. S. JENNINGS
Now, suppose that in this culture there existed no specimens
near this mean size, but only the larger and smaller specimens.
What would be the nature of the correlation? To determine
this we must omit all pairs containing medium specimens, and
compute the correlation only for those containing extreme in-
dividuals. If there were no correlation, no tendency for like to
mate with like, we should among these extreme individuals find
large specimens matched as often with small as with each other;
the correlation would be 0.
Let us consider as ‘medium’ specimens all those included in the
three groups above and the three below the group containing the
mean; that is the seven groups nearest the mean, in table 11;
these are marked off by the lines a, b, c, d. Then the extreme
specimens are those lying entirely outside these lines. There
are twenty-three pairs containing only such extreme individuals;
computing the correlation for these, we find it to have the high
value of 0.678 + 0.054. For the lot as a whole the correlation
is but 0.398 + 0.030. If in the same way we compute the cor-
relation for all pairs in which no extreme individuals are present
(the 87 pairs within the small square enclosed by the lines a, b,
c, din table 11) we find a still smaller coefficient, of but 0.104
==) 0/041.
To complete the picture, we may ask what the correlation is
when we select all medium individuals as principals, and compute
their correlation with their mates, whether the latter are ‘medium’
or ‘extreme,’ and do the same for all extreme individuals. We
find that there are 157 pairs belonging to the former group, 93
to the latter. The 157 ‘medium’ individuals are correlated with
their mates to the extent of 0.229 + 0.036; the 93 ‘extreme’
individuals are correlated with their mates to the extent of 0.439
+= (0.056.
Thus we find that there is a steady increase in the positive cor-
relation as we include more and more ‘extreme’ individuals, the
coefficient beginning at 0.104, and becoming successively 0.229,
0.398, 0.439 and 0.678. ‘This is exhibited in the diagram of fig. 14.
These relations are general. I have worked out the correla-
tion for the ‘extreme’ and ‘medium’ specimens for a number of
CONJUGATION IN PARAMECIUM 49
a
Or Nh Wh
| OO
A D
Fig. 14 Diagram showing the different degrees of correlation between members
of pairs in lot 1 (table 11), according as we examine the mating of individuals of
medium sizes or of those of extreme sizes. O-O, mean and line of no correlation;
B-D, line of complete positive correlation. 1, Correlation (0.104) for pairs con-
sisting exclusively of individuals of ‘medium’ size. 2, Correlation (0.229) when
we take all specimens of medium size as principals, and compare them with their
mates. 3, Correlation (0.398) whea all specimens, of all sizes, are included.
4, Correlation (0.439) when we take all specimens of ‘extreme’ size as principals
and compare them with their mates, of whatever size. 5, Correlation (0.678) for
pairs consisting exclusively of individuals of extreme sizes. (For the sizes con-
sidered ‘medium’ and ‘extreme’ see table 11).
the lots mentioned in tables 1 and 2. The results are shown in
table 12. Column 1 answers the question: What would be the
correlation if no extreme specimens existed? Column 2 gives
the correlation for the culture as a whole, while column 3 answers
the question: What would be the correlation if no specimens
of medium size existed? ;
As appears from the table, in every case the extreme individ-
uals show higher correlation than the ‘medium’ ones, as well as
higher correlation than the culture as a whole. In all cases save
THE JOURNAL Of EXPERIMENTAL ZOOLOGY, VOL. Il, NO. 1
50 H. S. JENNINGS
TABLE 12
Correlation between the members oj pairs (a) for extreme individuals only, as com-
pared with the correlation for (b) medium individuals only, and (c) for the entire
collections, including both medium and extreme specimens. (Column IV shows
what are arbitrarily considered ‘medium.’ specimens for the lot in question; the
extreme spectmens are those lying outside the limits designated medium)
I | II II IV
|
| - = | a
|
non MEDIUM PAIRS | ALL PAIRS EXTREME PAIRS
re j = —S= saa | SES CONSIDERED
um- | *“MEDIUM’
lars Gocticlent: of Nom, Correlation pia Correlation ‘ ;
(Wild culture)
l 87 0.104+0.051 | 180 0.398+0.030 | 20 0.678+0.054 198-212
2 45 0.320+0.064 | 142 0.268+0.057 | 29 0.518+0.092 15/-173
3 39 0.015+0.076 82 0.507+0.049 | 16 0.830+0.052 154-204
6 59 0.421+0.072 136 0.428+0.033 19 -0,6140.096 172-192
7 27 0.021+0.130 79 | 0.333+0.048 | 22 0.477+0.111 160-172
(Pure races) |
9 36 —0.'27+=0.111 | 125 | 0.132+0.042 | 34 | 0.176+0.112 144-156
16 49 0, 256+0.090) 150 0.507+0.029 | 41 | 0.843+0.022 120-136
19 WeexA 0.052+0.129 87 | 0.323+0.046 | 16 | 0.681+0.090 112-124
(Random mating)| | |
Au 45 0.051+0.100 | 105- | —0.108-+0.046 | 16 | —0.192+0.162 | 163-177
Cu 29° 0.187+0.121 | 101 | 0.045+0.047 | 27 | 0.0000.092 167-182
one, the correlation for the ‘medium’ individuals is less than that
for the culture as a whole. Further, on the whole, the greater
the correlation of the lot as a whole, the greater the difference
between the correlation of the ‘medium’ and the ‘extreme’ speci-
mens.
It may be well to note that this greater correlation of the ex-
tremes is by no means a necessary result of mere arrangement
in a correlation table. If the individuals are merely paired at
random, there is no significant correlation either in the culture
as a whole, or in the extreme pairs. This may be illustrated from
the random pairings made by Pearl (’07). Pearl wrote on sep-
arate slips of paper the lengths of all the individuals concerned
in the pairs; mixed these together, and drew out two at a time,
forming thus ‘random pairings.’ He did this for two lots, A
and C, containing respectively 105 and 101 pairs of conjugants.
I have worked out the correlation for the medium and extreme
specimens for the tables so formed (Pearl’s tables A 11 and C
CONJUGATION IN PARAMECIUM oil
11). The results are given in the last two rows of our table
12. As there appears, the extreme pairs do not show positive
correlation, any more than do the rest of the collection.
It is further to be noted that this higher correlation between
the extreme specimens than between the specimens of medium
size is not a necessary consequence of the existence of a consider-
able degree of positive correlation in the table as a whole,—though
it is doubtless a very common accompaniment of such positive
correlation. But it is easy to form tables showing a marked
degree of positive correlation, in which the correlation of the
extreme parts is not greater than that of the medium parts.
Why small differences between the individuals should not act
so precisely in determining correlation as do large differences
will be clear to anyone who considers carefully the process of
mating, as described on previous pages. The difficulty in pair-
ing caused by slight differences between the two individuals
concerned is readily overcome by slight curving, shifting, etc.,
while great differences are not so easily remedied. Hence speci-
mens differing much do not often unite, while those differing but
little unite readily. Thus the correlation tables may be expected
to exhibit many pairs in which the two members differ slightly ,—
and this of course prevents the correlation from being complete.
2. Different categories of pairs following different rules
A second condition that would result in incompleteness of cor-
relation would be the existence in the lot of different categories
of pairs, following different rules. A certain set might, taken by
themselves, give complete or nearly complete positive correla-
tion, while another set, following different rules of union, might
show little or no positive correlation, or even negative correla-
tion. The lot might then give as a whole bit a moderate degree
of positive correlation.
Is there any ground for suspecting the existence of such diverse
categories of pairs in our material?
Careful examination of the pairs shows that there is such
ground. The assumption on which is based the explanation
52 H. S. JENNINGS
of the existence of positive correlation is that the pairs in conju-
gation place their anterior tips in contact, so that in the pairs as
we find them the anterior ends of the two members should be
even, as in fig. 3, c, e, z, ete. Pearl (07, p. 267) notes that this
is on the whole approximately true in most pairs; he does not
give measurements on this point. But examination of a large
number of cases shows that (as Pearl further noted) the anterior
tips are not always even.
Now, if placing the anterior tips evenly together results (as it
should) in high positive correlation, then if in any cases the an-
terior tips are not placed evenly together; if the anterior tip of
one individual is placed some distance from the tip of the other
individual (as in fig. 15, b, c, e), then this would naturally result,
for such pairs, in less positive correlation, or in no correlation,
or even perhaps in negative correlation.
We might perhaps then expect to find at least two categories
of pairs, giving different results so far as correlation is concerned:
(1) those with anterior tips even; (2) those in which the anterior
tip of one individual projects beyond that of the other.
I have made an analysis of certain cultures with relation to
this matter, with the following results:
First, as we have before seen, observation shows that the two
members of a pair are by no means always equal, but that num-
bers of unequal pairs are found. A number of such are shown in
fig. 15. Cases of extreme inequality sometimes occur, but such
are rare. In one of the pairs of lot 2 (table 1), the anterior tip
of one individual extends forward thirty microns beyond that
of the other —that is, about one-fifth of the length of the latter.
Fig. 15, 6, shows a pair in which the smaller is less than three-
fourths the length of the larger.
Unevenness at the anterior ends. In four lots of conjugants I
undertook to measure the differences between the anterior tips
of the pairs. The measurements taken are necessarily somewhat
gross in comparison with the minute absolute amounts that
one individual projects beyond the other, but by using large
numbers we may get results that will be accurate enough to
indicate the real conditions. We may call the individual that
CONJUGATION IN PARAMECIUM 53
Fig. 15 Unequal pairs. a tod, pairs of the aurelia race c; e and /, pairs of cau-
datum; g and kh, pairs of the aurelia race C2. a, b, and c give each two views of
certain unequal pairs. X 333.
54 H. 8. JENNINGS
projects farther anteriorly A, the other 6; the measurements
then show, beside the total length of A and B, the amount that
A projects in front of B. One of the four lots thus analyzed
was a ‘wild’ culture, the others were pure races.
The wild culture examined was lot 2 (table 36); the unit of
measurement in this case was 2 microns. Pairs in which the
difference at the anterior ends was less than 1 micron were con-
sidered even. The results for lot 2 are given in table 13.
TABLE 13
Distance that one member (A) of a pair projects in front of the other (B), in a ‘wild’
culture (lot 2, table 36)
5
TOTAL NUMBER | MEAN LENGTH
OF ALL, IN
Ces MICRONS
Projection of A in mi-/ 0| 2, 4/ 6{| 8 | 10/ 12 | 14| 16 | 18| 30
CIOUSssasepislewiclecis ea | |
Number of cases......|61| 9 23/15/15|10| 4) 2/1] 1] 1 We Hie
Per cent of total
number of pairs.....|43.0) 6.3)16.2)10.6:10.6) 7.0) 2.8) 1.4) 0.7) 0.7) 0.7
Thus of the entire 142 pairs, 61 (or 483 per cent) were even at
the anterior tips, while 81 (57 per cent) were more or less un-
even. The amount of projection of A was rather slight; in 123
pairs, or 86.6 of all, it was less than 5 per cent of the mean length
of the individuals.
From pure races there were examined from this point of view
one lot from race C, and two lots from race g. These are both
races of aurelia, and of about the same size, the mean length of
the conjugants falling, in all three lots, between 118 and 124
microns. In these cases the unit of measurement was 4 microns
instead of 2. The facts are given in table 14.
In these cases we find respectively 33.3, 45.2 and 46 per cent
in which the anterior ends are even. The amount of difference
at the anterior end is again rather slight. A difference of 8
microns is about 6.5 per cent of the means length in these cases;
this difference is exceeded, in the three lots, respectively by 11.5,
1.2 and 8 per cent, of all.
CONJUGATION IN PARAMECIUM 55
TABLE 14
Amount that one member (A) of a pair projects beyond the other (B), in three lots
from pure races
|
NUM
TOTAL UMBER| MEAN LENGTH
OF PAIRS |
| |
Projection of A, in microns..... 0 4 8 12 16 20 |
Lot 17: Race C2 , | | |
Number of cases...........- 23 24 14 E 3 69 121.91
Per cent of all.....:........5. | 33.3 | 34.8/ 20.3) 7.2 4.3
Lot 18: Race g | | |
Number of cases 4 .+..| 38 37 8 1 84 123.37
Per cent of all... sine of 45.2.)/) 4450/1) -925)| 12
Lot 19: Race g | | | |
Number of cases.... ..-| 40 30 104} 56 1 87 118.28
Percent otall...............-| 46.0 | 34.5 | 11.5] 6.9 | 14
Now, from the theory as to the cause of the correlation, we
should expect to find that the individual A, projecting in front
of B, is as a rule larger than B. (We shall later show that the
length from anterior tip to mouth is so correlated with the entire
length that this must be true.) This being so, we should per-
haps expect that the individual A, which projects farthest for-
ward, should likewise project farthest backward, thus overlap-
ping B at both ends. (If however the two specimens merely
came together at random, and any parts of the oral surfaces
united, then there is no reason why the specimen projecting ante-
riorly should be larger, and as a rule the specimen that extended
farthest forward would not extend so far backward,—one speci-
men being merely displaced forward as a whole, with reference
to the other).
Examination of a number of cultures from this point of view
shows that as a rule it is true that the specimen extending far-
thest forward is the larger and likewise extends farthest back-
ward. The facts for five cultures are given in table 15.
As the table shows, the specimen A, projecting anteriorly, is
larger than B in from 83 to 91 per cent of all unequal pairs, while
it is smaller than B in but 2 to 11 per cent. Further, A projects
beyond B backward as well as forward in 51 to 67 per cent of
all, while B extends beyond A in the rear in but 10 to 28 per cent
of all.
56 H. S. JENNINGS
TABLE 15
Proportional number of cases in which the individual A, which projects in front, is
larger than B, and projects behind it as well as in front
A LARGER A SMALLER AL ae pete fea)
TOTAL | NUMBER | ; i
Lor NUMBER | OF PAIRS ar 3 —— =
| OF PAIRS | UNEQUAL | Abso- Abso- 5 Abso- Abso-
lute | cont | 'ae | cent | Inte | cont | Ite | dene
2 142 81 68 | 84.0 9 11.1 54 | 66.7 23 «| 28.4
7 79 «| 36 30) 83.3 5 13.9 | 26 | 72.2 6 16.7
17 69 | 46 40 87.0 2 4.3 | 28 | 60.8 8 17.4
18 84 46 41 89.1 1 2.2| 31 67.4 5 10.9
19 87 47 43 | 91.5 1 2.2)| 24 | 61d 5 10.6
Thus, where one member of a pair extends further forward
than the other, that member is usually larger. We should there-
fore expect that in pairs where one member extends further for-
ward than the other the difference in length between the two
members would be greater than in the case where the two are
even at the anterior end. We should further expect that the
difference in size between the two members would be greater,
the greater the amount that A projects anteriorly beyond B,
The facts with regard to these points are given in table 16.
The difference between the two members was taken in units
2 or 4 microns, in different lots; in the table the grouping is by
intervals of 4 microns. The number of pairs showing the larger
differences is of course small; on this account I have thought it
well to give the probable errors, as well as the number of pairs
in each case.
The table shows clearly that the difference between the mem-
bers of the pairs is greater in pairs in which one individual pro-
jects anteriorly in front of the other, and the greater the projec-
tion, the greater the difference. These things are by no means
matters of course; if it were merely necessary for the animals
to coalesce by any part of the oral surfaces that came in con-
tact, they would not be true.
Ou
“I
CONJUGATION IN PARAMECIUM
TABLE 16
Average difference in length, in microns, between the members of pairs, in relation to
the distance that one member (A) extends in front of the other (B)
Ze x |
Mand
= = ro) < 0 | 2-6 | 6-10 10-14 14 18 18-22
ogks
ZRz 5 (ANTERIOR
<0°0 Paine
B2&yz| ENDS EVEN)
2 ame
a
- on | oD
8 2 2 2 2 5 | 5
& = 2 | 3 z PS
la res a MEAN & MEAN a MEAN Pa MEAN | ~ | EXCESS
2 Leanne is eee oF) & (es OF | 5 eieaes OF iS Recess Bi cee 5
% p INTENGTH) 3 | LENGTH | § | LENGTH 8 | LENGTH &% | pencta | § Lenora &
45 a Aa ms a | a
alas 2 2 3 2 a | =
3 |o% 2 5 4 D 5 b
9 a c i Felice!
Biel fs 53 5 BO eeeahone
> 5 Bg Zz ag
ee aE 5 s | 8 ges
oe ies Zz 8 | 2 z |
2° g A | 17) |
5 Zz { | | |
61 AIAVL
Pa}VUN [US apy pa7]r4y
sainjjno aps ay? wouf sind 1of asoy, yyim paundwoo sp ‘uoynundas sajfo s.inoy awos
paqy squnbnluoo fo survd fo ssaquawu Lof yjbua) ur W01D)]aLL09 pup woUDLLDA fo sjuDj}sUo)
70 H. S. JENNINGS
obliqueness is very evident when the pairs are examined with a
binocular microscope; it is indicated in fig. 3, 7 and k. Asa
result of it, when the entire length of one member of a pair is
shown, the other will be a little foreshortened. Thus the measure-
ments taken will indicate that the two are less nearly equal than
they really are, and this of course will reduce the correlation.
In the case of the separated pairs, this difficulty is not met, since
the two individuals are measured separately.
c. Another cause, which might actually increase the measured
correlation after separation, would lie in the increase of size which
takes place after the separation of the pairs (see page 21). If
the interval of time since separation varies in the different pairs
(as is of course the case), then some will have increased in size
more than will others. But the interval since separation will
of course always be the same for the two members of a given pair.
The result will then be a series of pairs differing in size, giving
thus marked positive correlation between the members. It
seems probable that this is the chief ground for the actual increase
in correlation after the pairs separate.
Observation shows however that the relative sizes during union
are retained after separation (so far as this can be shown with-
out actual measurements of the living united pairs). In many
cases I noted that certain pairs, when first seen, were unusually
large or unusually small, or unequal. When they were meas-
ured, twelve to eighteen hours after separation, this was still
true.
In any case it is clear that the correlation in the united pairs
is not due to the temporary change of form during union, since
it persists after union ceases. The correlation is not due to the
equalization in mating.
Change in variability and in correlation of parts after separation.
In a preceding section we saw that the individuals of the united
pairs differ very markedly in certain respects from non-conjugants.
We have now an opportunity to determine whether these pecul-
iarities of the united conjugants are due to a differentiation of
these conjugant individuals, existing quite independently of the
act of union; or whether they are merely the result of the exist-
CONJUGATION IN PARAMECIUM ial
ence of the union, with its accompanying change of form. These
peculiarities of the conjugants were as follows:
a. The variability of the part anterior to the mouth is greatly
reduced, as compared with the case of the non-conjugants.
b. The correlation between the length anterior to the mouth,
and that posterior to the mouth, is greatly reduced in the united
conjugants.
ec. The correlation of the two members of a pair is much greater
for that part of the body anterior to the mouth than it is for the
total length of the body.
If these peculiarities are, as our previous examination had
seemed to indicate, merely the result of the changes during union,
then they should disappear after the pairs have separated and the
individuals have reassumed their usual form. We may therefore
get light on this matter by comparing these relations in the con-
jugants before separation and in those after separation. The
data for this comparison are given in tables 18 and 20. These
show the following:
1. The variability of the part anterior to the mouth. is not
ereatly changed after the conjugants separate (table 20, B).
In the wild culture of lot 22 the variability did indeed increase
after separation, from 5.60 to 7.28, but in the pure race (lot 24)
it remained substantially the same. Of course the separated
conjugants include no young specimens, while the non-conju-
gants (table 18) include both young andold. It is therefore not
to be expected that the variability will reach the same degree in
the former as in the latter. The data do not indicate that there
is any great change in the variability of this part, owing to the
mere act of conjugation.
2. The reduction of the correlation between length anterior
to the mouth and that posterior to the mouth quite disappears
after the cessation of union. In the united conjugants we find
in tables 18 and 20 coefficients of correlation between these parts
of but 0.261, 0.383 and 0.277, while in the separated conjugants
the correlations are 0.649 and 0.488 (table 20, G, 5). The same
lot that shows a coefficient of 0.277 while conjugating, has a ec-
efficient of 0.488 after conjugation has ceased. This is additional
2 H. S. JENNINGS
TABLE 20
Dimensions, proportions and correlations of conjugants after separation, as compared
with those for pairs still united. The two individuals in a pair are denominated A and
B. Compare table 19
(WILDCAUDATUMOFAUG.| WILD CAUDATUM OF | RACK k (AURELIA) OF | tees
3l-sepT. 1, 1910 SEPT. 21-22, 1910 SEPT. 11-12, 1910 Oman
Lor 22 Lor 23 Lor 24 19, 1910
| Conjugants 134 Pairs 134 Pairs | 178 pairs
102 pairs | after sep- ai eae, after separ-. ee ates after sep- | after sep-
/still united |aration, 276) > Tables ation | Tables aration | aration,
Tables Tables 60+ Tables
Table 55 | specimens FE
Table 56 | 22+23 | 57 and 58 | 5% 62, 64 |61 ¢3, 65, 66| 67-+68
| | |
(Mean......... 176. 14£0. 48/212. 46+0. 6¢|179. 80+0. 41200. 940.79 118. 92+0. 33/144. 87+0. 40/132. 96-+0. 43
pie { Stand. dev...| 10.08+0.34) 16.97+0.4¢| 10.42+0.29| 19.18+0.56 7.54£0.23} 9.590. 28) 12.07+0.31
aie (Coef. of var..| 5.72+0.19] 7.99+0.23) 5.80+0.16 9.540.283) 6.34+0.19] 6.62+0.19| 9.08+0.23
B. Length {Mean.........|111.02+0.44/128.64+0.38 71.33+0.20| 86.76+0.23
in front |S. dey...) 6.21+0.31] 9.37+0.27 | 4.58+0.14) 5.60+0.16
of mouth |Coef. of var.., 5.60+0.28) 7.28+0.21 | 6.42+0.20] 6.45+0.19
C. Length ee ee 66.13+0.51) 83.83+0.38 47.59+0.21| 58.10+0. 22)
behind 4Stand. dev...) 7.22+0.36| 9.34+0.2i 4.75+0.15| 5.43+0.16)
mouth | Coef. of var..) 10.91+0.56) 11.14+0.32 9.98+0.31) 9.35+0.28)
D. Proportion of entire
length lying in front of |
the mouth........ z 63.0% 60.6% 60.0% 59.9%
E. Mean.........| 29.21+0.18} 43.08+0.19 47.96+0.27) 29.020. 16) 45.970. 16)
Breadth 4 Stand. dev...) 2.66+0.12) 4.58+0.12 6.59+0.19 2.614£0.11) 3.89+0.11
Coef. of var..) 9.11+0.43) 10.63+0.31 13.75+0.41) 8.98+0.39 8.47+0.25,
F. Proportion of mean | |
breadth to mean length. 16.6% 20.4% 23.9% | 24.4% 31.7%
G. Coefficients of correla-) |
tion
1. Total length of
individual A with! |
same of B....... 10.359+0.0410.411+0.0340.245+0.0370.358+0.0360.210+0.041) 0.432+.033/0.231+0.034
2. Part before mouth) |
in A with same |
Thy ee recdanrenoe 0.400+0.034 (0.435+0.0350.370+0.036
3. Breadthof A with)
sameofB......... 0.356+0.034 0.295+0.038, 0.325+0.037
4. Partbeforemouth |
with total length |
(of same indi- |
vidual) 22csrtetecra 0,9120.007 '0.820+0.0140.886+0.009
5. Part before mouth
with part behind
mouth (of same
individual)....... | 0.6490 024
6. Total length with)
total breadth (of |
same individual) | 0.622+0.025) 0.648+0.024 (0.504+0.031
|
| 0.277+0.040/0.488+0.031
|
|
CONJUGATION IN PARAMECIUM to
proof of the fitting and change of form in the anterior region
during the union.
3. This same thing is still more clearly shown by the change
in our third point. In the two members, A and B, of the united
pairs, the correlation of the parts anterior to the mouth (that
of A with that of B) is double the correlation between the total
lengths of A and B. But after separation the correlation is prac-
tically the same for the two dimensions. ‘This is shown in table
21.
TABLE 21
Correlation of parts anterior to the mouth, as compared with correlation of total length,
in conjugants (1) during union, and (2) after separation
CORRELATION BETWEEN PARTS CORRELATION BETWEEN TOTAL
1. CONJUGANTS DURING UNION ANTERIOR TO MOUTH LENGTHS
Lot 7, table 18 0.840+0.016 0.333+0.048
Lot 18, table 18 0.626+0.031 0.323+0.046
Lot 24, table 20 0.435+0.035 0.210+0.036
2. consUGAN'S 12 HOURS AFTER
SEPARATION
Lot 22, table 20 0.400+0.034 0.411+0.034
Lot 24, table 20 0.432+0.033 0.370+0.036
From all these facts it is clear that the parts anterior to the
mouth are fitted to each other during union, giving high correla-
tion. After separation the change of form that brought about the
fitting disappears, so that the correlation for this part becomes
no greater than that for the body as a whole. The correlation
for the entire body however persists undiminished.
Correlation of pairs not due to equalization. These facts of
course demonstrate completely that the correlation in total length
is not due to the contraction, curving, stretching, etc., that takes
place in fitting one conjugant to the other, for this correlation
is fully as great after the change of form due to this fitting has
quite disappeared. There is clearly an assortative mating that
is quite independent of this fitting process, larger individuals
mating with larger, smaller with smaller.
74 H. S. JENNINGS
Another conceivable method of equalization is discussed by
Pearl (’07). He suggests that this might occur by a passage of
fluid from the larger to the smaller individual of the pair. He
points out that this would be a process requiring some time, and
that therefore, if this were the explanation of the correlation, the
latter should be higher in pairs that have been in conjugation
for some time, than in those in the early stages of the process.
By sorting out those in early stages of the process (as shown by
the nuclear conditions), and comparing them with those in later
stages, Pearl showed that this is not the case. Thesame thing
will be shown experimentally later in this paper; pairs taken in
the earliest stages of conjugation give as high a coefficient of cor-
relation as those which have been united several hours (see table
24 and the adjoining discussion).
Of course there is absolutely no indication from any source
whatever that any appreciable quantity of cytoplasm passes from
one member of the pair to the other. Thus the suggestion of
equalization by this means is one so entirely without foundation
as to hardly require for its disproof the evidence above given.
3. Change of size during union
A eause that might conceivably produce correlation between
the members of pairs is the following: The conjugants might
change in size in some typical way, either decreasing or increasing,
during the period in which union continues. Thus, if they de-
crease in size, the pairs at the beginning of the period of conju-
gation would be large; those near the end of the period would be
smaller, and intermediate ones would show intermediate sizes.
If we took then a collection containing paws in various stages of
this process, we should find a marked degree of correlation in
s1ze.
It is therefore important to determine whether such a typical
change of form occurs during union. For this purpose I tried
the following experiment. A wild culture of Paramecium cauda-
tum was placed, on the evening of September 20, 1910, under
conditions favorable for conjugation (many specimens in a shal-
CONJUGATION IN PARAMECIUM 75
low watch glass). At this time there were no conjugants in the
lot. Now, as Maupas (’89, p. 171) has noted, under such condi-
tions Paramecium eaudatum begins conjugation in the early
morning. At five o’clock the next morning rare scattering pairs
were found. Others were beginning to unite, so that as I watched
them the number of pairs increased rapidly. Beginning at six
a.m. I picked out about 200 pairs as quickly as possible; all these
were isolated, with no admixture of non-conjugants, before seven
TABLE 22
Correlation table for lengths of conjugants of lot 23 in
the earliest stages of conjugation. (Unit of measure-
ment, 4 microns)
40 41 42 43 44 45 46 47 48 49 50 51 52
38 1 1
39 | 2 1 1 4
40/1 1 1 3
41 1);2)2/1 2 8
' 42 3/4/2/2/2)1 1 15
43 GN 2) 2 7 V3) 4 16
44 2|)2/5 1 1 11
45 3) 4/213) 1 1 14
46 1 3/3 2 9
47 2)6)42) 1) 71 )1 2
48 1 1
49 1 1
3 1 | 7 |15 |11 14 111 117 | 8) 5,2) 1 95
a.m. At this hour I killed about one-half the lot; their measure-
ments are given in table 22. This includes pairs in only the early
period of union. The remainder were kept for five hours, till
noon; then these were killed; their dimensions are given in table
23. If there is any typical change of size during conjugation,
comparison of these two tables will show it, since in the latter
table the animals have been in conjugation five hours longer than
in the former. Furthermore, if such change of size is the cause
of the correlation, then a table comprising both setstogethershould
give a higher coefficient of correlation as well as a higher coefficient
of variation than either of the tables alone; and notably higher
76 H. S. JENNINGS
than the coefficients for the pairs in the early stages of union
(table 22). The data for the two lots separately and for both
together are given in table 24.
As table 24 shows, there was no change in size during the five
hours of union between the taking of the first and second sets.
TABLE 23
Correlation table for lengths of conjugants
of lot 23 in later stages of conjugation.
(Unit of measurement, 4 microns)
42 43 44 45 46 47 48 49 50 51
40 | | |i
41/2/11] 2
42|2/1 1
43 3
44 } 1/2
5
1
—
a
ro
me
45
47 | | |
48 |
49 oh a
}4}2/6/4\14/9/6/4]3]1 53
to
i
an
ry
WNOaIntwoar
TABLE 24
Constants of variation for the pairs of lot 23, after different periods of union
BER MEAN STANDARD COEFFICIENT OF erases
s a 2 , q - =
OF DEVIATION VARIATION IN LENGTH
Pairs in the earliest
stage of union:
F AM Sane scree site 95 179.79+0.53 10.79+0.37 6.00+0.210.303+0.044
Pairs that have been
united more than | |
five hours: 12 m.... 53 179.81+0.64 9.72+0.45 5.41+0.250.118+0.065
Pairs in early and late
stages of union taken
together (sum of the
two foregoing)..... 148 179.80+0.4) 10.42+0.29 5.80+0.160.245+0.037
CONJUGATION IN PARAMECIUM Ul
The mean lengths for the two sets are perhaps nearer together
than one would usually hope to get them for two samples of the
same material. Further, there is no increase of variation nor of
correlation when we unite the pairs that have been but a short
time in conjugation with those that have been united a long time.
In fact the coefficients for the two sets taken together are slightly
less than for the lot in the first stages of union, though the differ-
ences are merely those that might be expected when two samples
of the same material are examined. It is clear therefore that the
correlation is not due to any characteristic change in size as union
continues.
4. Differential contraction due to killing fluid
Another cause which might conceivably produce correlation
between the members is the following: A contraction or other
change of form due to the action of the killing fluid might vary
among the different pairs, owing to the impossibility of all coming
in contact with the same concentration of the fluid at the same
relative instant. If such differences occurred, they would always
be between different pairs, and not between two members of the
same pair, because both members of a pair keep together, and
would therefore be subjected to identical influences. Thus
we should have some pairs that were contracted and therefore
short; other longer; the result, when all were taken together,
would be to give a correlation in length between the members
of pairs.
Any change of form due to the killing fluid is evidently very
slight, if it oceurs at all; nevertheless, Paramecium does possess
a certain contractility and it seemed best to test carefully this
possibility. For a time indeed I was inclined to attach consider-
able importance to it. An opportunity for a test is given by the
experiments already detailed, in which the conjugants were meas-
ured after they had been separated at least twelve hours. After
the pairs have separated of course the two will no longer be sub-
jected to identical action of the killing fluid; all correlation should
therefore disappear, if it depends on this action. As a matter
78 H. S. JENNINGS
of fact, as we have seen, the correlation does not disappear, and
is not even lessened under these conditions.
However, in such experiments, the two members of a given
pair are usually killed together in a single minute drop—each pair
separately. It might still be maintained that there are accidental]
differences in the action of the killing fluids on the different
drops, causing different degrees of contraction in the different
pairs; this would therefore still give us correlation.
In order to test this, in a number of lots, I killed the two in-
dividuals of each pair separately, in separate drops. If the fac-
tor we are discussing plays any part, it would now act strongly
against correlation, since the two members of a pair would con-
tract diversely. There should be no correlation in such pairs,
if correlation is due to diversity in action of the killing fluid.
In certain Jots I had three sets of pairs: (1) those killed while
united; (2) the two members separated, but killed together: (3)
the two members separated and killed separately. Comparison
of these will enable us to set at rest absolutely the question of
the part played by any differential contraction due to the killing
fluid. The pertinent data are given in table 25.
As this table shows, the correlation persists even when the
separated members of pairs are killed separately. Thus the cor-
relation is certainly not due to any differential action of the kill-
ing fluid on the different pairs.
One peculiar case in the table requires mention. In lot 24,
we find that for the separated pairs in which the two members
were killed in the same drop, the coefficient of correlation (0.506)
is much greater than for the case where the two members were
killed in separate drops (0.230). This taken by itself would seem
to indicate that the killing of the two members together does have
a decided influence in increasing the correlation. But further
examination shows that there is no such indication. First, in
this same lot the correlation for the separated pairs killed sepa-
rately is greater than for the pairs still united. It is clear therefore
that the positive correlation of the latter is not due to the fact
that the two members of the pair were killed together; and this
is precisely the question we are putting to the test. Second, in
79
PARAMECIUM
CONJUGATION IN
'ye0'0- 182 0.8z 0+80°6
‘8400 20Z 0£E 0-+6T 6
‘gh0'0-F 82 028 006'8
|
|
1e°0-L0°ZI SFO 96 CET) SLI
Fr'OFS0E ICI OFEH SEI] 39 [16 | AoyuIUdes ports !poywaudo:
Zh OFFL IT 09086 'iSI| L9 (28
-*- 1994080} pal] ‘payeredag *e
Shes qpuevjowng ‘o}
>| &%
(4) eIfeny
: | | 99
180'0SZE 0880 0 ZEF OGL 0FZO°9 [820696 OF OFL8 FT T9+09/FET “-°(9-+9) payeandas [TV *P]
960'0F 082 OIG OFL0'S [IE OFTF'L [FF OFL6 SFI) 19 oo. Aleyeredes pertry tpayuredag °O| | z
ZE0'0F 90S OTE OFOL'L |SF OFOT IL FO OFES EFT) 09 69. |" "10430709 party! paresedag -q{ | "°
TF0'0*0IZ O\6T OFFE'9 [8Z'OFFG'L [S206 BIT] GF \CoT payun ins *8) |
| | (4) BIEN
88° OFG6Z 0 980 0F8GE08Z OFF G6 [99 OFT GT 62 OF FG 000 Lg pet | Ajoyeredas poly tpaqyuandas “q) | ez
LEO OF SEZ O9T OF08'S [60 0FZF OT Th O08 GLT/E2-+esFI | payan ag “ef |
| | | UIN}EPNB! PITAL
$20 0-998 OFEO'OFIIF OLS 0F66'L [6F'OFLGE°OT 69 OF 9F GIG 9 [SET | 1099709 PATIL *poyeredag q| | __
TF0' 0698 OT OFZL'S |FE°0F80 OI SF OFFI OLT) SF (COT poyrunypyssyaemntuog “ef | °°
| | al UINgepNed! PTT |
| |
ee lee oie we ied E
. | | a q |
Q4eSe/Sesse|) BSS | wae | Bis Ve
waebe |egeea| See | 2o8 pee
B8a|~scea|] Ses S25 g Fs
wo | Bed 4 z i) |
meee | AS&e | Ree Aad x om
aaa Z| ie all o4 ° a | 3 |
ron] 88e8 | 44 ag fc ”
gage) gaze/ ssc | 4 Fs I |
He ally see nae g emis |
pouyuasagip ybnowy} wornjasoa fo woyonpoud ajqussod D 389}
“pungf Surry 247 fo woxon
so WIAVL
07 syupbnluos fo uosr.wwdwo)y
8O H. S. JENNINGS
the case of the separated pairs killed together, not only is the
coefficient of correlation greater, but the variation is likewise
much greater than in the case of those killed separately. In the
former the coefficient of variation is 7.76, and the range of varia-
tion was from 88 to 168 microns, while in the latter the coefficient
of variation is but 5.07, and the range was from 128 to 164 microns.
Now this great difference in variation cannot possibly be due to
the fact that the two members of the pairs were killed together
in one case, separately in the other; there is absolutely nothing
in this procedure to change the variation. Clearly, the differ-
ence between the two lots is purely accidental. Among those
taken for killing the two members together, it happened in this
particular case that all the extreme specimens fell, leaving the
medium individuals for the other lot. Both samples are rather
small, so that this is not particularly surprising. Now, we have
seen on previous pages that collections containing only specimens
of medium size give a lower degree of correlation than do col-
lections containing extreme specimens. This is clearly the ex-
planation of this peculiarity. Such fluctuations in the coefficient
of correlation due to the accidents of random sampling are not
particularly rare. In one case, random matings of the compo-
nent dimensions of a series of 125 pairs gave me a negative coefh-
cient of correlation of somewhat beyond —0.500, though there
was absolutely nothing in the process of drawing the numbers from
a hat that would tend to induce correlation of any sort.
6. Local or temporal differentiations in the culture
An important possible cause of the production of corre’ation
in the members of pairs is set forth by Pearl (’07) as follows:
It might be maintained that since at different pomts in the culture
and at different times the environment no doubt differs slightly, there
would be a corresponding local differentiation of the Paramecia in each
local culture unit. Then, even though the pairing were quite at random
in each locality, yet if the records for several such localities were mixed,
a spurious homogamie correlation would arise (p. 256).
CONJUGATION IN PARAMECIUM 81
Larger pairs would come from one part of the culture, smaller
from another, giving of course a positive coefficient of correlation.
This result would be much accentuated if the pairs were taken
from the cultures at different times, as was actually the case in
the study made by Pearl.
Pearl attempts to meet this by taking the two non- conjugant
individuals nearest to each pair, and hence from the same environ-
ment, pairing these, and making a table of the results. Such
pairs give no positive correlation.
I am not convinced that Pearl’s procedure fully meets this
difficulty. In the non-conjugant population we have to deal
with the great variability due to growth. Of the two single
individuals lying nearest a pair, one might be old and large, the
other young and small. The difference might be much greater
than any differential effect of the environment. In the case
of the conjugants, on the other hand, this difference in growth
plays little or no part. That is, in the non-conjugants we have
an important source of variation that is independent of the en-
vironment, while in the conjugants we have not. It would appeas
likely therefore that local environmental differentiations would
have much more effect on the correlation of conjugants than on
that of non-conjugants.
In the case of the conjugants studied in the present paper this
source of error was completely excluded for most of the lots ex-
amined, through the method by which they were taken. On the
evening before the day the conjugants were desired a great num-
ber of individuals, none of whom were conjugating, were taken
from the large culture and placed together in a small watch glass.
They were thoroughly stirred and mixed in the process. The
watch glass contained only water, besides the infusoria. On the
following day the animals were conjugating in multitudes. There
had been absolutely no opportunity for local differentiations such
as might give origin to correlation, yet collections so made showed
the same degree of correlation as did others taken from larger
vessels. In the case of collections of conjugants taken from large
cultures, I removed individuals from only one small region and
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, No. 1
82 H. S. JENNINGS
all at the same time, so that the results of local differentiation
were avoided in these cases also.
The fact that the coefficients of correlation for my samples
are notably less than they were for Pearl’s (see page 38) perhaps
indicates that these local differentiations may have played some
part in Fearl’s.results. As Pearl himself remarks ‘‘the samples
used in this (his) work were taken in just such a way as would
make most pronounced any spurious correlation due to local
differentiation resulting from time or place factors. Small
samples, a drop or two of culture fluid, were taken from different
parts of the culture at intervals of time”’ (l.c., p. 256). But the
results of the present paper show that the effects of this were
limited to increasing to a certain extent the degree of correlation
observed. They do not account for the existence of the correla-
tion.
6. Other suggested causes
Pearl raises the question ‘“whether these correlations represent
any true assortative pairing or merely arise because conjugation
goes on within a limited, differentiated portion of the population,
which portion, as has been shown above, is much less variable
than the non-conjugant population”’ (07, p. 254).
It is not apparent how this last mentioned condition could
produce correlation, since the latter is not affected in any way by
the characteristics of that part of the culture which is not conju-
gating, and low variation has no tendency to cause correlation.
However, Pearl demonstrates fully that this is not the cause of
the correlation. He makes random pairings between the measure-
ments of the conjugating individuals, and shows that these ex-
hibit no correlation, as they must do if the explanation just
mentioned is correct.
7. Correlation in breadth
In the present paper I have dealt mainly with correlation of
the two conjugants in length. It is here that the most important
relations show themselves. Furthermore, as Pear] pointed out,
CONJUGATION IN PARAMECIUM 83
the measurement of breadth in conjugating pairs is very inaccur-
ate, owing to the flattening that takes place as conjugation occurs.
Also, the ridge forming the right side of the boundary of the oral
groove of one specimen fits into the oral groove of the other, pro-
ducing an interlocking which makes it almost impossible to deter-
mine correctly the breadths of the two individuals.
On these accounts I have not thought it worth while to deal
with the breadth in any such full way as I have dealt with the
length. Anyone who is interested in a careful analysis of the
breadth relations of the two conjugants will find it in the paper
of Pearl (07). I shall take up only a few points not brought out
by Pearl.
a. Flattening at the time of conjugation. Pearl states that at
the time of conjugation there is frequently a dorso-ventral flat-
tening of the two individuals, such as would result from pressing
the two together; he gives however no measurements on this
point. I have endeavored to obtain some precise data on the
matter in certain cases.
In the small race ¢ (aurelia) (lot 9 of table 2), I measured the
amount of flattening in fifty pairs. This was done as follows:
The breadth was first measured as the two members of the pair
lie side by side, in the usual position. This gives the dorso-
ventral breadth. Then the animals were turned till one lies
directly above the other; they were kept in this position by plac-
ing them in a jelly made from quince seeds in water. In this
position the lateral breadth (at right angles to the dorso-ventral
breadth) was measured. In almost all cases the lateral breadth
was notably greater than the dorso-ventral breadth, showing that
the animals were distinctly flattened. The two breadths are
shown for these one-hundred conjugants in table 26.
I made the same measurements also for fifty non-conjugants
of this same culture. These were likewise flattened in the same
way, indicating that the flattening was not due to the pressing
of the two conjugants together, but that it existed in this culture
independently of conjugation. The dorso-ventral and lateral
breadths are given for the non-conjugants in table 27. The con-
jugants are flattened a little more than the non-conjugants;
84. H. S. JENNINGS
TABLE 26 TABLE 27
Lateral and dorso-ventral breadths (or Lateral and dorso-ventral breadths (or
thicknesses) of 100 conjugants of lot thicknesses) of 50 non-conjugants of
9 (race c). (Unit, 4 microns) lot 9 (race c)
Lateral Lateral
7|8| 9 (10 /11 (12 113 7 | 819 (10 |11 12 113
3 6| |3]1 4 4 1 1
Se oa eGR aati 15 5/1 1 9
S 8/1) 2 12 11) 3)1 30 3S 6/1 9 3
2 9 15 20/4] | 39 ss. 7a) Oulena eam liak 8
aS 10 Op) ely als) dit Sg 21613 ll
1 1 1 2 9 2/5 1 8
i | Pai S 10 4 4 8
2 | 8 119 30 34 | 6 | 1 {100 11 poe ete ar bal (C33
3 12 2 2
13 i) 3
2 50
4/3 116|9{|6 10
in the former the mean dorso-ventral breadth is 83.4 per cent of
the mean lateral breadth; in the latter it is 86.5 per cent.
Thus in certain conjugating cultures there is a notable differ-
ence between the dorso-ventral and lateral breadths, even in
non-conjugants. It is important however to bring out clearly
the fact that in most cultures of Paramecium, whether conjugating
or not, there is no such flattening. I have made measurements in
the same way of several other cultures, and have examined many
more, and in no other case have I found any such marked flatten-
ing. In this culture of the race c, the animals were thin and
showed every appearance of being in process of starvation. I
believe that this is the explanation of their exceptional dorso-
ventral thinness. In most cultures the animals are nearly or
quite circular in cross-section.
b. Correlation in breadth in pairs after separation. Pearl gives
the coefficients of correlation in breadth for three lots of pairs
that are still united. It seems worth while to add here the breadth
correlations for certain lots measured after the pairs have sep-
arated. In such eases there is no such difficulty in making accur-
ate measurements of breadth as we find in the case of pairs still
CONJUGATION IN PARAMECIUM : 85
united. The correlation in breadth for three such separated lots
is given in table 20 (G, 3). The values for the coefficients (0.356,
0.295 and 0.325) are not far from those found by Pearl] for unsep-
arated pairs (0.218, 0.342, and 0.349).
ce. Correlation in breadth not due to equalization. It is worth
pointing out that the correlation in breadth of the two members
of a pair could not be produced by that equalization of the pairs
through stretching, contraction, curving, ete., which we have
discussed so fully in the foregoing pages. For as the leneth of
the two was made more nearly equal by this process, the breadths
would inevitably become more unequal. The shorter member,
stretching to match the longer, would become more slender;
the longer specimen, contracting to match the shorter, would
become broader, thus increasing the differences already existing
(existing as a result of the fact that the longer specimens are al-
ready broader than the shorter specimens, a fact demonstrated
by Pearl (’07) and the present author (08). Thus the existence
of marked correlation in breadth between the members of pairs
is in itself a demonstration that the correlation is not due to the
equalizing during union; a demonstration made on other grounds
elsewhere in this paper (pp. 65-74).
&. Historical and comparative
Pearl (07) was apparently the first to notice the assortative
mating in any infusoria, and his study of the matter was more
thorough than any other made previously to the present paper.
We have dealt so fully with his work in the body of this paper that
we need not dwell further on it here.
Collin (09) observed that in the conjugation of Anoplophrya
branchiarum, a parasitic ciliate living in the blood of Gammarus,
the two members of a pair areusually of nearly equal size, although
the different pairs differ much in size. He thinks that there is
real assortative mating, larger individuals mating with larger,
smaller with smaller, as in Paramecium. However, a part of
the greater similarity of the two members of a pair is in Anoplo-
phrya due to the fact that as conjugation progresses the individ-
S6 H. S. JENNINGS
uals of the culture become smaller. Hence the first individuals
that conjugate form large pairs; later ones form smaller pairs,
still later ones still smaller pairs,—(although the size of a given
pair does not change during conjugation). This does not, Collin
thinks, account fully for the similarity in size of the two members
of a pair, for even among the individuals that conjugate at any
given period one finds much greater resemblance between the
two members of a given pair, than between members of diverse
pairs. Collin does not give measurements.
Enriques (08) studied conjugation in Chilodon uncinatus, with
relation to the problem of assortative mating. He found that a
change in form takes place during conjugation, by which the
left member of the pair becomes shorter than the right; further,
the evidence indicated that the left-hand member was even before
conjugation somewhat smaller than the right-hand one. This
latter relation would result from the fact that in the process of
conjugation, owing to certain peculiarities of form in Chilodon,
‘the larger individual always becomes the right-hand member,
the smaller one the left-hand member.
Assortative mating was studied with relation to the question
whether the two members of a pair tend to be of the same size or
not. That is, if the member A diverges from the mean of all,
does the other individual B Jikewise diverge in the same direction
from the same mean? Enriques studied this question by deter-
mining the mean difference between the two individuals of the
pairs, and comparing this with the mean difference in length when
matings are made at random. In ease of assortative mating in
the sense above defined, the former value should be less than the
latter. Enriques found that in most of his (rather small) samples
it was not less; so that there was no indication of assortative
mating in the sense defined. But in samples taken toward the
end of an epidemic of conjugation, the members of pairs were
more alike than were individuals taken at random, so that a
degree of correlation is indicated. For two of his lots Enriques
worked out the coefficient of correlation; for the early sample
the coefficient was zero; for the later one it was 0.4. Thus in
the later periods of an epidemic there is an actual correlation in
CONJUGATION IN PARAMECIUM 87
size between members of pairs. Enriques explains this difference
between the early and the late stages of the epidemic in the fol-
lowing way: In the late stages of the epidemic many of the pairs
are formed (wholly or partly) of individuals that have already
conjugated once in the same epidemic. This fact is determined
by studying the nuclear conditions of the pairs; some individuals
are found to be undergoing the changes consequent on a previous
conjugation. Now, Enriques found in Chilodon (as we have
found in Paramecium) that the ex-conjugants increase in size.
Hence those that are conjugating for the second time in any epi-
demic are larger than those that are conjugating for the first
time. Further, Enriques found (by studying the nuclear con-
ditions) that for some reason these ex-conjugants are likely to
conjugate together. Thus we get certain paws, consisting of
ex-conjugants, in which both members are large; other pairs,
consisting of individuals that have not before conjugated, in
which both members are small. This of course results in the pro-
duction of some considerable degree of positive correlation when
the entire lot is examined.
Evidently, the question of main interest is: Why do the large
ex-conjugants tend to conjugate together on the one hand; the
small individuals on the other? There is clearly assortative mat-
ing of some kind. If (as appears probable) it takes place on the
basis of relative size, as in Paramecium, then the interpretation
of the facts in Chilodon would be the following: In the early
stages of the epidemic the differences in size among the individ-
uals are not great enough to affect the mating, which therefore
takes place at random. But in the later period of the epidemic,
owing to the appearance of the large ex-conjugants, the differences
in size become so great as to prevent the union of the largest and
smallest individuals. Hence assortative mating occurs, giving
rise to a certain degree of correlation between members of pairs.
Enriques raises the question whether re-conjugation among
larger ex-conjugants may not be the cause of the correlation in
Paramecium. It is quite clear that this is not the case. (1)
Pearl (’07) shows that the correlation exists when one includes
only individuals showing early stages in the nuclear processcs
88 H. S. JENNINGS
attendant on conjugation. (2) We have shown above (p. 75)
that when pas are taken in the very first stages of anepidemic,
before it has been in progress more than two or three hours, there
is still correlation in size in the members of pairs. It may be
noted that there appears to be no evidence that in Paramecium
re-conjugation ever takes place among ex-conjugants.
Attention should be called to the fact that the method used
by Enriques does not furnish a test for another possib'e kind of
assortative mating. In Chilodon, the right and left members
of the paw are morphologically differentiated and the right one
is usually larger. Now, it is possible that assortative mating
may so oceur that when the right member is larger than usual,
the left member must also be larger than usual, though it need
not be so large as the right one. That is, when the right hand
member is above the average size for right-hand members, the
left member may be above the average size for left-hand members.
This might still be true, even though the average size for right-
hand members were considerably greater than that for left-hand
members. Ii this were the case, then the average difference
between the two members of a pair would be greater than the
average difference between members of random matings, and yet
this would not be evidence against the kind of assortative mating
we have mentioned. This kind of assortative mating could be
tested by determining the coefficient of correlation between all]
the right-hand members (entered as X) on the one hand, and all
the left-hand members (as Y) on the other, each set being referred
to its own mean (as in the usual computation of correlation),
instead of to the mean of all. A parallel case for this is set forth
for Paramecium on p. 58 of the present paper. Unfortu-
nately, Enriques has not given us the correlation tables for the
actual measurements of his pairs, so that the existence of this
kind of correlation can not now be tested for Chilodon.
ies)
ve)
CONJUGATION IN PARAMECIUM
9. Conclusions on assortative mating
The results of our examination of the causes of the correlation
between the members of pairs (an examination that can perhaps
fairly be designated exhaustive) 1s to confirm Pearl’s conclusion
that there is actual assortative mating in Paramecium—larger
individuals mating with larger, smaller with smaller. This shows
its effect in three main categories:
1. When the two species caudatum and aurelia are conjugating
at the same time in the same culture, they do not inter-cross;
caudatum conjugates only with caudatum, aurelia with aurelia.
Since the two species differ in size, this gives very high coefficients
of correlation (0.940, in length). The only reason that the co-
efficient is not actually 1.000 is the fact of incomplete correlation
within each of the two component species.
2. When races of different size are present, as is usually the
case in ‘wild’ cultures even of a single species, the members of
the larger races tend to conjugate together, on the one hand; of
the smaller races on the other. The correlation (in length) thus
arising is less than when two species are present; it averages about
0.380, in the wild cultures of the present paper.
3. When members of but a single race are present (all descended
from a single individual), there is still a notable correlation, larger
individuals mating with larger; smaller individuals with smaller.
But this process of assortment is considerably less accurate than
when diverse races are present; for pure races the coefficient of
correlation averages but about 0.250, as against 0.380 for mixt-
ures of races, and 0.940 for mixtures of species.
II. CONSEQUENCES OF THE DIFFERENTIATION OF THE CONJU-
GANTS AND OF THEIR ASSORTATIVE MATING
“ What effects in the further history of the organisms result
from the occurrence of conjugation? What are the consequences
in inheritance, of the decreased size and variability of the conju-
gants, as compared with the non-conjugants? What con-
sequences follow from the assortative mating of the conjugants?
90 H. S. JENNINGS
CONSEQUENCES OF THE DECREASED SIZE AND LESSENED
VARIABILITY
1. Are the extreme specimens excluded from the new generation?
Does the decreased size and lessened variability of the conju-
gants indicate that extreme individuals of the race are excluded
from taking part in the new generation, conjugation tending thus
to maintain the average racial type unchanged?
As we have seen, the non-conjugant population includes many
individuals that are larger, and some that are smaller than any
of the conjugants. Are these smaller and larger individuals
excluded from participation in the further development of the
race?
a. As to the smaller individuals, examination of the compar-
ative tables (tables 34 and 35 of the Appendix) will show that
the non-conjugant population usually contains but few individ-
uals smaller than the conjugants, its main differentiation lying
in the opposite direction. Now, we have already shown that these
small non-conjugants are merely young specimens, that have not
yet reached the size for conjugation. A few hours will of course
remedy this. The small individuals are then not excluded from
conjugation.
b. The main difference between conjugants and the general
population is that the latter contains many specimens much
larger than the conjugants. Are these larger individuals excluded
from conjugation, and so from the further progress of the race?
Cr are these larger individuals merely temporarily differentiated,
their progeny becoming conjugants later?
This matter was tested in a number of cases by removing from
a conjugating culture a considerable number of the large non-
conjugants, each much larger than any of the conjugants, and
keeping them under conditions favorable for conjugation.
Thus, January 30, 1908, I removed from the conjugating cul-
ture of lot 6 (table 34) one hundred of the large non-conjugants.
On the next day fifty of these were killed and measured; they
had a mean length of 211.52 = 1.99 microns; a mean breadth of
CONJUGATION IN PARAMECIUM 91
62.48 +1.17. They had multiplied a little during the night, so
that this is really less than the original size. The conjugants
at the same time had a mean length of but 181.49 += 0.54; mean
breadth 48.11 + 0.87. The remaining large non-conjugants were
allowed to multiply, and on February 24 their progeny were found
to be conjugating. The conjugants which they gave were of
sensibly the same size as those of the culture as a whole. The
mean length, from 21 pairs, was 179.91 = 0.88 microns; mean
breadth 48.19 + 0.51 (as compared with 181.49 by 48.11 for the
culture as a whole).
Thus in this case the large non-conjugants were by no means
excluded from conjugation and the farther development of the
race. They represented merely a temporary stage, not as yet
prepared for conjugation, but ready to enter upon it after a few
fissions.
Similarly, from a wild culture NV there were isolated on March
21, 1908, one hundred of the largest non-conjugants, and a con-
siderable number of pairs of conjugants. Half of each were killed
at once; the fifty large non-conjugants showed mean dimensions
of 224.16 by 53.36 microns, while the sixty-six conjugants meas-
urec. 139.94 by 38.55 in mean length and breadth. On the fol-
lowing day numerous conjugants were found among the progeny
of the remaining fifty non-conjugants. The large individuals
were thus but temporarily excluded from conjugation.
A similar experiment was tried with the small pure race c
(aurelia). On September 27, 1907, I isolated from a conjugating
culture ten of the larger non-conjugants (larger than any conju-
gants); also five pairs of conjugants. These were cultivated side
by side, under the same conditions. On September 30 there
were conjugants among the progeny of the ten large non-conju-
gants. (On the following day there were likewise conjugants
among the progeny of the pairs.)
Again, from the pure race Nf, I isolated, March 31, ten of the
largest non-conjugants; on the following day there were conju-
gants among these.
Thus, it is clear that the large non-conjugants may later divide,
take on the size characteristic of conjugants, and themselves
conjugate.
92 H. S. JENNINGS
2. Relative size of progeny of conjugants and non-conjugants;
changes in size due to conjugation
Do the progeny of the large non-conjugants differ in size and
variability from the progeny of the (much smaller) conjugants?
The answer to this question is bound up with that to another,
so that the two will be considerea together. This other question
Is:
Are there characteristic changes in size resulting from con-
jugation, so that arace Just before conjugation is smaller or larger,
or otherwise characteristically different from the same race just
after conjngation?
The answer to this question will tell us whether there are char-
acteristic changes in form and size in the different periods of the
reproductive cycle (from conjugation to conjugation), for just
before conjugation the animals are at the end of the cycle; just
after they are at the beginning of it.
To answer these questions experiments were carried out on a
number of different cultures. The plan of experimentation
was in most cases as follows: From a conjugating culture a
sample of non-conjugants was removed; also a sample of conju-
gating pairs. Part of each of these was killed and measured,
thus giving the initial size for each. Then each set was allowed
to multiply farther, under rigidly identical conditions for the
two sets. Samples of the progeny of each were killed at intervals,
and compared as to size. The method of work was variedin
different cases, in ways that will be set forth.
This plan of work, simple as it sounds, in reality presents great
difficulties in its execution. Each individual must be kept sepa-
rate, in order that we may know that there has been no conju-
gation save in the case of the original pairs; and also in order that
we may know to what generation of the progeny a given individ-
ual belongs. Owing to these and other difficulties, it is not
possible to obtain under these absolutely controlled conditions
large numbers for comparison of the progeny of conjugants and
non-conjugants, though the numbers given in the following are
sufficient for answering the main questions.
6
CONJUGATION IN PARAMECIUM: 93
It will be well to examine first the results in the case of pure
lines or races, then to take up ‘wild’ cultures.
a. Comparison of progeny of conjugants and non-conjugants in
pure races. The first comparison was made between the progeny
of conjugants and non-conjugants all descended from the single
individual Nfs. In the progeny of this individual conjugations
occurred March 31 and April 4, 1908. In each case a random
sample of the conjugants and non-conjugants was killed and
measured; the relative sizes and variabilities are given in the first
four lines of table 28. The non-conjugants were as usual con-
siderably longer, and more variable than the conjugants. On
March 31 there were picked out five pairs of conjugants and ten
of the largest non-conjugants. In this case the individuals were
not kept isolated, but the ten conjugants were kept together in
one watch-glass, the ten large non-conjugants in another, under
identical conditions. Ten days later (April 10) samples of each
of these were measured, giving the results shown in the fifth and
sixth rows of table 28. The progeny of the pairs were now a
little larger than the progeny of the non-conjugants.
The progeny of the very smallest pair observed on March 31
was kept separate from the remainder. On April 20 its progeny
were compared with those from the ten large non-conjugants
taken on the same day. The results are given in the last two rows
of table 28. The size of the two sets was now very nearly the
same (those derived from the small pair were a trifle longer and
thinner). The variability of the two sets was now practically
identical.
All together this experiment shows the following facts:
1. The progeny both of the small conjugants and of the large
non-conjugants increased somewhat in size beyond the original
mean for the non-conjugants. This was probably due to the
nutritive conditions.
2. The progeny of conjugants became fully equal in size both
to the original non-conjugants and to the progeny of these non-
conjugants. They were in fact a little larger than the progeny
of the non-conjugants, and the difference appears to be significant
in comparison with the probable error (at least on April 10).
JENNINGS
Ss.
He
94
{ | | j
Ze OFC8 EF) 98-98 GE°OFOL'L [So°0FS0 ZI SL: OFSF 9ST\Z6I-ZET | OT | [dy
| | | jo syursnfuod-u0ou
| | | | JIB] OT Jo Auaso1g go, ‘0% [dy 8
| |
GE OFFO' GH 9S-9E [$6 0F89'L [LE OF9E ZI [0S OFG8 O9T|SSI-OZT | 80T | # [dy jo ayed
| | | Tews | Jo Auatorg 80, ‘0g WAdy 2
| | | }
82°000'FF | ¥9-8z |00°0F L001 [28°006 FT OTT 46 LETIGLI-FOT | 99 | Te 2eIy |
| | | jo sjyuvsnfuoo-uou |
| | jesivl OT Jo Auasorg 90, ‘OT ey 9
£9008 GF | 0298 |12°0FS9°CI 60 TOs SFI|ZLI-OcE | 19 | Te “aeN 30 sured | |
| | | | yews ¢ jo Auas014 80, ‘OT ney | c
£9 OF 01 GE FF-8Z |LO TFS CI 9S TST SI [0 GFIO LFISSI-O8I | 18 ““syuBsnfu0d-u0N|g0, ‘F [4dy) &
0€ 080 TE | OF-8Z FE OFIN'S CFOF99D 9 OFS ZETSFI-GII Og ““squvsn{u0D|g0, ‘F Tudy ¢
SOF 00 FE | FSS GP OFFS GNF L [6 0F6S FHI SII-CEL | FE | “syUBNfuoD-aoyl\r9, ‘Te “~AVIN| &
09°0*2S'SE | FE-FZ 6 OFL0F [TF OFESE 9C 06 OETSFI-FEI | ZF squvsntuoD gd, ‘Te “Ve 1
| i}
z | = Q a | z a) 2 | | ©
| 2 i} Pa | sel ed | >
eo eee ee BE B |SER2| Ee 2
= BOG Reta! fio 4 g459| 25
c 154 zs 2 2 agz | #8
= egieng ob ee || g | Hg}
a HZ OZ | x “ig il 10,
> < 24 0 fe Rit i)
o > ies] = Z >| nD
te} mt iS} < ) bh
a fn - 7 Boe 5
| | | 7 a
(prjainv) *f nN Jonprarpur ajburs ay? wos fjjoULb1L0 paarsap
ny ‘fuabosd way) fo pun syunbnCuos-uou puv spunbnluos fo fiqyiqnisa pun azis aaynjay
86 ATAVL
CONJUGATION IN PARAMECIUM 95
3. The variability of the progeny of the conjugants increased
and became sensibly the same as that of the progeny of the non-
conjugants.
From September 16 to September 26 a similar experiment was
carried out with the race of Paramecium aurelia that I have called
C,. The measurements for conjugants and non-conjugants of
September 16 are shown in the first two rows of table 29; the non-
conjugants are considerably larger. At this time four pairs of
conjugants were placed in each of twelve watch-glasses; eight of
the large non-conjugants in each of twelve others; the two sets were
kept under identical conditions and treated in the same way.
The table gives the sizes of the progeny at certain later dates.
As the table shows, the progeny of the conjugants were larger
than those of the non-conjugants on September 18; less so on
September 23, and smaller on September 26. Possibly the only
conclusion clearly justified is that the progeny of the conjugants
are not on the whole smaller than the progeny of the (larger)
non-conjugants.
From September 25 to 29, 1908, a similar experiment was under-
taken with the aurelia race g, which is closely similar to C,. In
this case the second and third generations of the progeny were
TABLE 29
Relative size of conjugants and non-conjugants and of their progeny from the pure
line Cy (aurelia)
C
Sa ies a "
a a a
a |s # < 5
& > 5 > S|
o | 84m z fs 2
x one a = =
82) eaoo Ms i) :
2 m2) ou Zz on a Zz
& 2a | 3498 < Zoek <
< pa} sa a < i So Be I)
=
a
mmoreanonawtwon
amo aoKr nana
|
|
|
|
|
yyouaTy
112
fo Uy
‘squpbnluos-wou ‘xy fspupbnluoo
SsUoLoU F “Wuauiaunspaut
‘(saunqjna yons fo saunjxiu Lo ‘DyainD
‘
0)
fo saonu aind fo sainqjna Ayurvw) oe¢-g $10) fo sjunbnlu0s-uow puv syunbnluoa fo syybuaT
st ATAV.L
CONJUGATION
TABLE 36
IN PARAMECIUM
113
Measurements (in units of two microns) of the 142 pairs of lot 2 (wild culture), classi-
fied according to the distance that A projects in front of B. The first column gives
a classification of the various lengths of the projecting individual A; the other col-
umns give the lengths of their mates B for each pair, under the different anterior pro-
jections of A. Thus, there are eight pairs with A 76 units long; their mates B
measure respectively 75, 73, 74, 76, 76, 72, 74, 74 units; in the first four the anterior
tips of A and B are even; in the others A projecis in front of B 2,4and7 units,
respectively
LENGTH OF MATES, B, WHEN THE DISTANCE A PROJECTS BEYOND B, 1s
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, NO. 1
| 8
|
0 1 O93
|
70
72, 73
(2)23, 74, (2)76 | 72 74
|
(2)68, 72, (2)74, (3)76 73, 74, 76 | 78 72
78, 79 |
1(2)76 | 72
78, (2)79, 80 |76.79 | 80
76, 81 78 74
78, 79, 82 79 | 80 | 75, 76, 78 | 78, 81
74, 76, 78, 79 77, 83 77 69, 74, 84
75, 85 81 /79, 81 79
| 75, 79, (2)83, 84, 85, (3)86 | 77 85 76,83 | 85
77, 86 83 79
76, (2)83, (2)84 88 | 82 87 83
85, 86 86
87, 88 |80,88 | 79
84 79 82
84, 90 78, 87, 92 |
85 82
82, (2)93 76
92
90 78 /79
83 179
| 88 |
78
om
xan
oe
6
76
| 7
a
72
114
Lot 8 (wild culture).
43 44 45 46 47
EH: 5:
TABLE 37
48, 49 50 5
bo oo be
]
at
1
2\1
1 2
S28 [2
V4 3
1/2/1
2\3
5|5|5 |11 |11 22 |12
TABLE 38
JENNINGS
bo
Length of A by length of B
52| 53 54 55 56 57 58 59
1
|
| 1
1
1
1
3 | 2 1
Lot 4 (wild culture). Length of A by length
of B
31 32 33 34 35 36 37 38 39 40
29 1 1
30 1 1 2
Sie 22 3
32 eal 3 5
33 1 3/2 1OHi7,
34 D211 5
35 D3) 3 1/9
36 1p a 5
37 ihe 14
38 1
1 | 4|3]2)10|8)3)6 3 |42
_
45
49
CONJUGATION IN PARAMECIUM
TABLE 39
35 36 37 38 39 40 41 42 43 44 45 46 47 48
1
Pali age
1
aly Pal i}
ial Later tel 1
1H a ef Lt if 1
2) 4 2
Dee Ds Sats) S25 hel
SHSy wor Ieee) lay LN pe a a
463. \eo.2") a
1)/4)4/4)3
BE | Pet) a) a Es}
CT Cs ha Da fe |
1); 4/2
4
| 1| 3) 2 18 12 |14 16 16 20 13 |13
Lot 6 (wild culture). Length of A by length of B
49 50 51 52 53) 54
115
116 H. S. JENNINGS
TABLE 40
Lot 7 (wild culture). Measurements, length from anterior end to mouth, and total
length; classified with relation to the distance that A projects in front of B. (The
lengths ‘‘To mouth’’ are not repeated when successive pairs are the same in this
respect.) (1 unit = 4 microns)
ANTERIOR ENDS EVEN ANTERIOR ENDS EVEN PROJECTION 1 PROJECTION 1
To Mouth Total To Mouth Total To Mouth Total To Mouth Total
B A B A B A B
B FGM bats) A B|A|B A |
25 | 25 | 37 | 36 46 | 42|| 25] 24 | 39 | 37 51 45
40 | 37 | 28] 28] 41] 41 | 41 38 Sa ee
PROJECTION 2
42 40 42 40 42 36 i z
26 | 26 | 39 | 38 | 42/41] 26] 25] 40/39] 25/ 23] 38] 36
| 40 | 39 | | 43 | 40 40 | 41 | 27| 25 | 39 | 36
| 41 40 | 43 | 42 27 26 38 | 39 41 | 41
42 41 43 | 42 40 | 38 | 28| 26] 44 | 38
42 | 42 43 | 42 43 | 37 44 38
42 | 42 43 | 42 44} 41) 29 27 | 42| 43
43 39 43 | 43 | 44 | 42 44 40
43. 41 44 | 42 44 | 42 48 44
44 | 37 48 | 44 | 28) 27] 41 | 39 48 | 45
45 | 42] 29] 29) 43 | 42 41 | 42 ms
PROJECTION 4
46 41 46 | 42 | 42 | 41 =
46 | 41 46 43 42| 43 26| 22] 44 | 32
27 | 27 | 41 | 41 46 44 44 | 41
42 41 47 | 42 44 | 42
42| 41) 30| 30) 46) 45 44 | 42
43 40 47 | 42 44 42
43 41 48 | 45 48 | 40
44 | 41 49 16) 29} 28 | 43 | 39
44| 43) 31] 31) 47 | 44 | 46 44
CONJUGATION IN PARAMECIUM nlalz¢
TABLE 41
Lot 7 (wild culture). Non-conjugants; length before
mouth with length behind mouth
Length before mouth
24 25 26 27 28 29 30 31 32 33 34 35
SSS
12h il Via el ae ht 2
See] Eh as ene Ba 17
S14) 2h pa | 4
2 15)| si 2h) ti 11 9
m3 16/1/1/2/3)5)1/2/3)1 19
S17 3) |1)2/4/2/4)1 17
'$ 18 0) ed li Une |e 2a 13
2 19 | (2piideel ed rales | |6
220 Ay ze 3 elk
a) 21 | ome i | pa eA ale FOG
22 he 3 | | |3
23 | 1 | 1 | 2
mist aiediete pee laid
2 {10 | 8/11 [12 [10 | 9 [15 14) 1] 1 | 1 94
| { cere | Keel
TABLE 42
Lot 8. Length of A by length of B
31 32 33 34 35 36 37 38
Se ae
30 2 | | 2
Sine Belin | ee 3
32 let 1
33 | 1 2 Teo 7 1 | 9
34 (BT ete seh a 5
35 | Pye eal 5
36 | ipl he ga faale 9
|
: | aa)
1\/7/2)4)6)5|1)] 1 (27
118
H. 8S. JENNINGS
TABLE 43
Lot 9 (race c). Length of A by length of B
33 34 35 36 37 38 (39 140 41 42 |43 lea las
1 1 2
}1|t 2 | 4
1 | | 4
2 eal Ot) 13 10
Lael ae | Swale alartun Marg
2/3/4/5| |3 17
217|3|5|6 23
oH Aes: 3 20
2/2/5)5|6/1 | 1 | 22
i be es 7
Pa 3s I 8
1 | 1
L 1
2/3 6/7 (18 18 12 23 13 16/3/31 (125
TABLE 44
Lot 10 (race k). Length of A by
length of B
30 31 32 33 34 35 36
28 Ne a 1
29 iy p34) al Al
30°), 1: C2 yeoTl ae) 5
31 Esl) Aa | Lae 7
32 Ppl aba at (ete 5
33 2 | 2
34 | 1 1
CONJUGATION IN PARAMECIUM
TABLE 45
Lot 11 (race k). Length of A by
length of B
LS)
oo
29 30 31 32 33 34
24 1 1
25 | 1 1
26 | 1 1
27 | ty} 2 2
PAST ul Pena (ea Ee PG an 5
29 al
30 1 Le) (2
31 1 1
bo
2 243 1 | 14
TABLE 46
Lot 12 (race k). Length of A by
length of B
28 29 30 31 32 33 34 35
26 1 1 2
DNe2 ste) (len a 5
28 Sesame sys 10
29 TOV t RAs OE VE, er 9
30 31316 12
31 3771915 24
32 Sulu a2 8
33 4} 2) ).2))\ 8
3 | 3] 8 |14 22 116 110 | 2 | 78
119
120
H. S. JENNINGS
TABLE 47
Lot 18 (race k). Length of A by length of B
126 [27 28 [29 /30 |31 32 33 34 35 36 37 38 39
Ate aa Ay Teta a eal
23 | 1 | 1
24 |
25 | J | 1
26 | Pe ela 1 | | 5
7 [at | il fe ea | 5
28 | 1132) 1S Sa" st 11
29 | AOA 10) Gey Ze lea | 1¢
30 1/12|6/3/6 1| ¥| 30
310 3.) 4 cba Syl eels alee
32 | 5/5|4/3|5/3/2|] | 27
33 | 4\/4/6/2/2/2) | 20
34] | }5/3)4/2] | 1] 35
Bt) 04 1j1lalal 4
36 W2elcoul oI 5
37 | | ja 1
110 i i 5 17 17 25 36 19 20 14 6 | 2 168
TABLE 48
Lot 16 (race C2). Length of A by length of B
26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
| | ees (oma eat |
= | [ | heals lal rte
25 | 1 | | | } 1
26 | |_| | | Wess ie 1
27) ns ey | 1 | 7
28 1 Ve WS atrouiea fete | 15
29 | | lslzisl2lsialaiiia | | 28
30 | 3/5) 4/4/2) | | | 18
31 | SR Oe RS Sa Oa a
32 | Ss anes aad | 20
33 | Wpeliai felts 534 3 6
34 }5 73/2/11 11
35 | Hee eeu salen 10
36 | | Hee lari Nay 7 8
37 | | (sya) | 4
| | | | | | t
11] 1/4 [13 a5 {14 (19 |17 |17 |12 16 [13 | 7 3/1 180
CONJUGATION IN PARAMECIUM " 121
TABLE 49
Measurements of length for the 69 pairs of lot 17 (race C2), classified according to the
distance that A projects in front of B. Unit, 4 microns. (Compare the explanation
of table 36)
LENGTH OF MATES, B, WHEN THE DISTANCE A PROJECTS BEYOND B 1s
A es ae =
0 1 | 2 3 4
26 | 24 22
2 27, 28 |
28 28 27 | 25
29 29 24, 25, 28, 29, | 27
30
30 26, 28, 30, 32 | 28 25, 27, 28
ol 26, 30 (2)31 27, 28
32 (2)31, 30 (3)29, 20, 31 31
33 24, 30, 31, (2)32) 30, 31, (2)32 27, (2)28, 30 27, (3)28, 30
33
34 (2)32, 34 33
35 (2)33, 35 33, 34 29 26
36 34
37 | 28
Total num-
ber......| 28 24 /14 5 3
TABLE 50
Measurements of length for the 84 pairs of lot 18 (race g), classified according to the
distance that A projects in front of B. Unit of measurement 4 microns. (Com-
pare the explanation of table 36)
LENGTHS OF MATES, B, WHEN THE DISTANCE A PROJECTS BEYOND Bis
A — - oo
oO 1 4 2, 3
28 27
29 (2)28, 29 27, 28
30 28, (4)29, (4)30 | 27, 28, 29, (2)30, 32 27
31 (2)29, (6)30, (3)31 | 28, 29, (2)30, 31
32 (2)31, 32 27, 28, (5)29, (3)30, 31, 32 | 27, 28, 29, 30
33 30, (2)31, (2)32 26, 29, 31, (3)32 30 29
34 (2)33, (2)34 30, (2)31, 33
35 32, 33 (2)30
36. 31
37
38 33
Total num-
ber 202. 38 37 8 1
TABLE 51
Lot 9 (race g). Measurements of conjugants in length from anterior end to mouth,
and total, classified according to the distance that A projects in front of B. (The
lengths ‘‘to mouth” are not repeated when successive pairs are the same in this re-
spect.) Unit of measurement, 4 microns. (Thus, in the first pair, the distance
from anterior end to mouth is 64 microns in both individuals, the total length of A
is 100 microns, that of B, 96 microns)
ANTERIOR ENDS EVEN PROJECTION 1 PROJECTION 2
To Mouth Total To Mouth Total To Mouth | Total
A B A B A B A B A B | A B
16. |) 16) | -25° | "24 | a7 | 16.270 26 | 17 | 15 | 290 | 26
28 | 26 | 27 | 26 18 | 16 | 208 eu 2i7
ge || alge eee Ih Beye 2 7auli 26 19 | 17 | 28° | 2
| OT Oy 7 | 27 | 29 | 95
| 28 27 e2Sianlee 30 | 29 | 98
| 98 | 28 | 18 |.17 | 27 | 26 | "84." ) 729
18 18 27 27 | 2433 27. | | 34 30
27 | 27 | 28; 27 || 20 | 18 | 32 , 27
2 8ia1n026 29-28 34 29
PA) || yy | 208 | o8m |) 2 | 219" 836 | 29
28) | 30) | 30 | 30 ~
29 | 28 | 31 29 | PROJECTION 3
29 |. 30 | 19 | 18 | 28 | 27 |—
29") #30 99 || 25 || 17° | 14 | 29 | 24
BY Il) ey 20), |) 227 19ealGmleesl | 28
30 | 28 | ZOU 28a POM Taniese: | 28
10} | eel 290, 298) SOTA ola Seeca, | 28
29 | 29 31 | 29 35 29
29 | 30 30a 228 23 ele OOM NSS! | sol
30 | 29 SO es28 a, lee ee ee
30 30 32 29 PROJECTION 5
31 | 29 32 | 30 |—————
Sie COMME 2ONN |e 1O OOUN Su ooeueeti: N35 || 26
| 31 | 30 29) || 129 |
31 | 30 S2anle 28
Bile |) Sel Hero Qala
20a 20 lecstn || 38 33 | 28 |
32 29 33 31 || |
3230 34 | 29 | |
32 | 30 35 29 |
SOMO Meme o1e | 200 1933) 32 il
3380
oe | S10 |
33 | 33
340] 939
Oe 21533) 130
35 30
124 H. S. JENNINGS
TABLE 52
Lot 19 (race g). Non-conjugants; length behind
mouth with length before mouth
Length behind mouth
9 |10 /11 /12 |13 |14 |15 16 |17 |18 19 [20
yaya abjal | | 2
aw
Se |
2 15/ | |
2 16/1) 1 2
= 17 ji 3/3/1] | 8
FS WD BD ile Nat 10
S19) |1] ft)3}3 18
= 20; | |4/2 7 3 15
91 | | 1) ae 4) (l/s Pata | | 19
22 1} 2])1/6)2)2 1} wea ils
PEI al Naud [ee 1 12
24 | Dil ala 55 2 Tt eeal| tel
25 Tae lid) 4
26 1 | 1 | 2
| 2] 8 [12 (24 j23 22 14 | 7 | 2/2/11] 1 1118
CONJUGATION IN PARAMECIUM
TABLE 32
Lot 20; conjugants yrom mixture of races Cz
andi. Length of A with length of B
] |
29 30 31 32 33 34 35 36 37 38 39
RPnNN Ne
i
—
—
rFPoWn ere
em
NOONAN AWWH FE
bo
eS woe be
e
—
1]4\/3i9lon7i|6|7\|4\1/1]| 6
TABLE 54
Lot 21; conjugants from mixture
of races Ly (caudatum) and k
(aurelia). Length of A with
length of B
28 29 30 31 |32 33 34
27:15 1 | 6
28)1 3) 1 eal se id
29 3 loa 4 ite 2 13
30 5|5)3)1 14
31 reeled |i 5
32 Hat | a 2
33 2 2
126 H. S. JENNINGS
TABLE 55
Lot 22; 102 pairs of unseparated conjugants; measurements of length to mouth,
total length, and breadth, for A and B of each pair. (In some of the pairs
only the total length was measured, as the table shows.) Unit, 4 microns
TO MOUTH TOTAL LENGTH BREADTH TO MOUTH TOTAL LENGTH BREADTH
A B A B A B A B A B A B
26 25 44 45 7.5| 7.51) 30 28 45 42 8 9
26 26 42 42 i is 47 44 foal
43 43 8 8 30 29 47 43 Said
27 25 43 43 ub 6.5 49 48 7.5 8
44 42 7 6.5 31 28 49 44 7.5: \7
27 26 39 40 6:5 | 7 31 29 47 45 tC
43 40 cf 7 32 29 48 44
46 43 8 7 32 30 47 43 fhe AALS
27 27 43 43 ef vi 33 26 48 42 8 6
43 42 7.5| 6.5 39 38 fe lif
44 42 8 6.5 43 42 6.5 6
44 43 7.5] 8 43 43 8 7
44 44 7 6.5 45 42 Tan
45 44 8.5| 8 45 43 87
46 46 7 tf 47 45 8 7.5
46 46 7 8 49 44 a |
28 25 47 44 a 6 51 45 te ANT,
28 26 41 2 ih ia =
42 44 7 8 Only total length measured
43 42 theaye | wie) iol. = ae
28 Q7 45 44 7 7 eet A LENGTHS OF MATES, B.
46 41 8 6 ——$<$—
46 44 7 6.5 t 41 38, (3)40
47 43 8 iq 3 42 39, 40, 41
28 28 46 43 8.5 | 6 7 43 39, 40, (3)42, (2)43
29 26 A 44 8 7.5 6 4 41, (2)42, 48, (2)44
45 41 7 7 10 45 40, (2)41, 42, (2)48,
29 27 46 41 if 7 (2)44, (2)45
46 43 6.5 | 7 8 46 38, (3)42, (2)43,
29 28 43 39 7 7 (3)45
45 43 8 7 6 47 42, 44, (2)45, (2)46
46 46 7 7.5 2 48 44, 45
46 47 7 8.5 2 50 44, 48
47 47 8 10 1 52 49
50 43 9 7.5
29 29 49 45 8.5 | 8
CONJUGATION IN PARAMECIUM 127
TABLE 56
Lot 22; 188 pairs of conjugants about 12 hours after separation. Measurements of A and
B of each pair, for length to mouth, total length, and breadth. Unit, 4 microns
| |
es | ela |e pleeie | fle} eye)
B Sr eee 5 elt ee 8 | § Biel ett eas
cS) atts < g Biel ai) lilt eR lees Pie eels =
a Pac a 5 ga = 5 Ba | @ B 54 che
2 | a ) \ a a } es | & | «@ } BR & a
alp\a B/A|B| A B|A|B\A| Bi Ale lA B\AB A) B\A|B, 4\B
| : if fi ia || | | Viner ie | iri {|(Gaeal | | ‘ i e=
25 | 24) 41| 39] 9.510 | 32| 30| 54| 57 | | 8.5) 33 | 33 52 | 54 [12 9.5 | 35 | 32) 58) 50 12 {10.5
27/27/45) 46 10.511 | | 53 48 11.5 10.5, | 53 | 58 10.512.5) 58 | 53 11 11.5
28 | 26 | 46 | 43 | 9.5) 8.5] 54 | 48 |11.5)10 54 53 10.511 | 5856 11 1
29 | 27| 46 | 40 8.5) 9 54/49 |10 | 9.21) 34/29 | 54/46 11 | 9 |} 60 | 52 14 12
29 | 29) 45| 49 10 |10 | 32) 31 | 49 | 49 |10.5) 10) | 57 | 48 |12.9| 9.5) 60.) 53 42.511
30 | 28 | 48 | 46 11 | 9 | 51 | 50| 9 salle 54 | 52 |11.5]11.5) | 60 54 11 11
50 | 45 11 | 8.5) 52] 49 |10 | 9.5] 56 | 46/11 | 9 || 35 | 33 | 57 | 55 12 (11
| | 50] 48 ]11 [10 | 52 | 50 /10.5)10.5) 34 | 31 | 54 | 52 /10.5)10 60 | 55 10.5115
30 | 30) 52 53 9.511 | 52| 51 11.511 | 54] 52/11 10 | | 61/58 11 | 9.5
31/26 48/4319 | 9 55) 51 11,510 | 155 48 11.5) 9 62 | 58 12.510 5
| | 52] 42 |10 | 9 57 | 52 |10 | 9.5 | 95 | 57 12.512 | 35 | 34 52 59 11 A
BL 27 | 82 49 [12 /12 | 32} 32 | 51 | 52 /10 |10 | 57 | 49 /13.5|10 || | 5955 11 10.5
31 | 28 | 54) 5010 | 9.5) | 52 | 52/12 11.5) | 57 | 52 |11.5/10.5]/ — | 59 | 57/11/11
|__| 49 | 48 |10.5) 9 ] 53] 52 |t0 jlo | 58 | 52 10.5 10.5 | 60} 55 12 1
31 | 29 | 51 | 48 110 | 9 53) 52 |l1 |9 | 58 | 52/12 |10.5|| 35 | 35 | 55 | 60 \10 13
j | 51 | 47 {11.5} 9.5) 53) 55 11 13 | | 58} 5312 j11 | | 57 | 56 10.5)10.5
50 | 47 10.510 | 53 | 56 |11 [12 | 159] 52 (13. [a1 | 60 58 11 11
31/30 50 | 48 11.511.5 "| 55 | 53 11 |10 | 34 | 32 | 54 | 52 11.5) 9 |) 36 | 31 | 61 | 52 11 10.5
|53!50/9 |9 | 33! 28) 56/45 !12 [10.5 54 | 52 12 [10 |) 35 | 32 | 59! 54 |11.5111
| | 51 | 47 {10.5} 9 1 33] 29| 49] 49 jst lat | 56 50 11.510 6l 54 13 11.5
| |54/53\10 Ho |s3}31|s4|srfio fio | | | 56} 54/12 [0.5] 36 33 | 57 | 55 {13 |t1
} | 52!) 52:).9° H11 54/51 10 (10 | 57 | 52 10 10 | 58 | 53 12/14
| 51 46 10.511 | 54) 51 (11 10 | B8)|/63 10.5, 9 | 36 | 34) 57 55 12 10.5
| | 53} 49 [11 [11.5 55| 52 |12 [12 | 34| 33 | 52 | 56 j10.5|11 58 | 56 12 {11.5
| | 5250} 9.5)11 | 57 | 54 |10.5)11 54 | 52 [11 | 9.5) 59 | 57 |15 |12
31 | 31 | 50) 51 | 9.5/10 | 33 | 32 | 50/50 |10 | 9 (55 55 11 10 60 | 58 113.512
52 | 50 10.511 | | 52| 55 9.5 | 9.5) | 56/53 9.510 | 36 35 58 36 12 2
52| 54/9 [11 | 53 | 52 |11.5/11 | [56] 57/10 [12 || | 59 | 59 |12 [11.5
| | 54] 52 |11 [10 | (54/54 11 10.5 | 57 | 52 /10.510 || 36 | 36 57 | 59 112 [13
32 | 28 | 57 | 48 [11.5) 9 || | 54] 54/11 [11 | 34 | 34 | 56 | 56 [12. 5|12. 5} 59 | 62 |12.5|13
32 | 29 | 51 44 /10.5) 9 54/55 | 9 |10.5 58 | 58 12 |13 || 38 | 30| 60 | 50/13 {11
| 52| 47 10 |10 | | 55 | 53 12° |14 | 59 | 58 |12.5/11.5| 38 | 34 | 62 | 55 [10.5] 9.5
52 49 10.511 |55/53/9 jlo | 35/27/57] 46 \12 |10 | 40/ 31] 65 52 |13.5\11
53 | 45 10 | 9 |} | 55 | 54 |13 |11 | 35 | 30) 56) 50 11.511 |
| 53 | 47/| 9.5) 9 58 | 51/11 [10
vetlt| lelelelolelsleleielmenrels|zielolr1 1/2 I
nN
N
S. JENNINGS
ao
Lon ma OM es
4 a
aA
Onn
HH.
AMMNAMHOONMANAAAHARMRANDKRM AS
a al al Sol
cl
ol
nN
al
£9 79. 19 09 6S 8g) 2g “od oo v6 ee ZS 1G OS 6F SF Lp fie + oF iF €F cP TF OF
|
suolaul ¥ iuyQ “s4ind fos Suaquiaue on} ay} fo yj6uay ur uoynjasio; Uu0y
-pundas sajfo sinoy gy jnoqn (sdoup aywundas ur) pappry syupbnluog *(ainz]na ppm) EZ 107
io)
ol 46 @IAVL
CONJUGATION IN PARAMECIUM 129
TABLE 58
Lot 23 (wild culture). Correlation in breadth of A with B in pairs
about 12 hours after separation, the individuals killed in separate
drops. Unit, 2 microns
= —————— = *
16 17 \18 [19 20 21 22 23 |24 25 26 27 28 29 '30 31 32
16 1 1 2
17 | 150 1 3
18 1) 1 She2 7
19 | 3 3 | 1 1 8:5
20 IS ele (RGA el,| <4 ye es 115
21 ne lea 5) 2/1 1 {11
22 AM ANITON E21 5.2 | 2 2 1 | 28
23 HOBO By 2.) 0 | 1 1| 14
24 Aantal 3 | dy 1 18
25 ABN Tp) 181k | 10
26 al) eit ez! | %
27 2) 11] | 3
28 Sale | 1] 5
29 | 1) 2
30 | | | Foie Sih Pen aan Ae
e = ; | | * eS ee a isl
1; 1111/3! 210) 8 29 14 22 6 16 TAT 6 134
rd
|
TABLE 59
Lot 24 (race k). Unseparated pairs, length
of A by length of B. Unit, 4 microns
26 27 28 29 30 31 32 33 34 35 36 |
25 1a 2
26 | 1 1 | 1 3
27 2/3/4/3)2 het 15
28 3/12;5)4)1/38 1 29
29 8 /11;}6)4/2)1 32
30 Reda Pea fn ol sees 2|3 | 29
31 Gree 2 9
32 2 2
33 1 1
1| 3 | 8 [24 \28 |25 113-/12 | 21 3 | 3 [122
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, NO. 1
130. H. S. JENNINGS
TABLE 60
Lot 24 (racek). Separated conjugants, both members of
a pair killed in the same drop of fluid. Length of A
by length of B
9 30 31 32 33 34 35 36 387 38 39 40 41 \42
9 |
P aks
|
|
rh
©
Now oe
Nr owe Hw
bo
eee Se
RPWONANWWNW hd tv
.
CONJUGATION IN PARAMECIUM 131
TABLE 61
Lot 24 (race k). Separated pairs,
the two members of each pair
killed in separate drops. Length
of A by length of B
34 35 36 37 38 39 40 41
|
awe
wow w :
RPNwWe
—
NAN RS oon
— * =
2/64/14 14/ 9/3) 3 [6s
|
TABLE 62
Lot 24 (race k). Unseparated
pairs; length to mouth in A by
same in B
16 |17 |18 |19 20 '21 |22
1o}-3 | 2] 1 | 6
16 | 2 |10 | 6 1 19
17 10 17 | 4/3 34
18 27 15 16) 1} 1) 50
19 3) 2/1 11
20 ibe id 2
Length Before Mouth.
H.’S. JENNINGS
TABLE 63
Lot 24 (race k). Conjugants about 12
hours after separation; length before
| | | | 7 {
jt? 18 119 20 21 22 23 24 25
ai aha 1
15 |
16 1 1
17 lima Pe 1
18 1|2 3
19. 3 1 ale (ped fens
20 9/4/5)/5/1/)11)18
21 | 11 33) 9}1 54
ap) 17 (13'|-8.} 11.39
23 4|5)1110
24 i eI
1 1,6 18 56 31 17. 4 184
TABLE 64
Lot 24 (race k). Unseparated conju-
gants; length behind mouth by length
in front of mouth, in the same in-
dividual
Length Behind Mouth.
8 | 9 10 [11 |12 |13 /14 |15 |16
Eg a i ees 6
16, }) , |S, L039) Nae 225.
17 |} A |S 25s 2 56
18 2/12 25 40 /16/4/2! (101
19 4/6 |15|8|4/1 38
20 | 1 EST 2} | 15
21 a) (te 3
22 1 1
| 1| 4 [23 56 95 |45 |13 | 6 1 (244
Length Before Mouth
TABLE 65
Lot 24 (race k). Conjugants about 12 hours after
separation; length behind mouth with length before
mouth in the same individual
Length Behind Mouth
6|7|8 9 10 11 12 13 14 15 16 17 18
TU SS ality | 1 1
15
16} 1 | i
17 |1 1 2
18 }2) 1) 3
19 2 SS aN ta 6
20 | | 4 Bye G. [Gs fa | 24
21 | 2 /18 j21 22 | 9 | 72
22 | 1| {15 30 129 |15 |} 4) 1) 95
23 | 2 | 6 13 |16 4 41
24 | 1| 1/2\10/4)1 19
25 Helles 4
1 | (1:13 43 67 83 47 8 5 268
TABLE 66
Lot 24 (racek). Conjugants about 12 hours
after separation; breadth of A by breadth
of B. Unit, 2 microns
25 26 27 28 |
19 20 21 22 23 24
1441] | 1
15 | | |
16
Ta\ete| he i
TSA Mn et ah LGM ley ha 1
1OM S| | Pag tae Po 1
20) |2)1)7)/2)/2)1)83 18
21 11/6 4) 11
22 20 17|2|6 54
23 1/15|1)4/1/ 2) 24
24 4|3|6 ml ales
25 Tele 1) 4
26 ie 1
27
| Telimel
134
H. 8. JENNINGS
TABLE 67
Lot 25 (race k; aurelia). Correlation table for lengths of
conjugants about 12 hours after separation, the two mem-
bers of any pair killed together, in the same drop
| | {
| ) |
27 \28 29 30 31 32 33 34 35 36 37 38 39 40 41
159 bate fsa | | 3
26 | 1 1
27 1 1 2
28 1 2 3
29 ie) a peeiee | al | 5
30 | Bele | 3) | 2 l 12
31 1) age Jee fe es Ws gs 13
32 | | P| 2) 4)-24).34 as) aide a 17
33 | | A} 2) 2°) 1 earl ga aie 14
34 | Bt | 412m | | 10
35 | 29. 1/5
36 | | 1h Len aL | 2
1}1/3]1) 3] 8 |15 14] 8 13 to) 5 | 2/2) 1 | 87
TABLE 68
Lot 25 (race k; aurelia). Correlation table for lengths of con-
jugants about 12 hours after separation, the two members of
any pair killed in separate drops
27 28 29 30 31 32/33 34 35 [36 37 38 39 40 41 42
25 | 1 1 |) Aiea 3
26 | | | 1 1
ia Wat tr ay | 2
28 nen | 2
29 1] |2 | 3
30 | ee ta a Pe 3 2 | 1 10
31 | 123} 0 ayes | 1 1 12
32 [2)7)riajair)eya 21
33 | Ala) 32) | 1 14
34 | 1 512,144 10
35 (tal eal 1 7
36 ee yab es 1 4
37 | 1 1 2
—— ta <= — —'l = < =
1 | S0een | 4 1|9|8 (13 12 1211 |9 5 | 3 1 | 91
THE REPRODUCTION OF PARAMAECIUM AURELIA
IN A ‘CONSTANT’ CULTURE MEDIUM OF
BEEF EXTRACT
LORANDE LOSS WOODRUFF anp GEORGE ALFRED BAITSELL
Sheffield Biological Laboratory, Yale University
TWO FIGURES
Previous work! with pedigree cultures of Paramaecium aurelia
and Paramaecium caudatum has apparently shown that the life
history of these forms, when bred continuously on infusions of
hay made up exactly the same from day to day, tends to run in
a cycle which terminates with the death of the culture. Previous
work has also shown that Paramaecium aurelia’? may be bred
indefinitely on a culture medium which is frequently varied.
In view of these results the following question suggests itself:
Is the longevity of Paramaecium on a ‘varied environment’
dependent upon intrinsic stimuli from the frequent changes of
the medium, or is a ‘constant’ medium of hay infusion unfavor-
able because it lacks some elements which are essential for the
continued existence of this protozoén?
To test this point it is necessary to find, if possible, a suitable
‘constant’ culture medium which contains all the elements which
the organism demands, and to determine its effect on the vitality
of Paramaecium when subjected to it for a considerable length
of time. If such a suitable medium is secured on which para-
maecia will live indefinitely, it is apparent that the possible con-
tinual daily stimulation afforded by ‘varied’ culture media is
1 Calkins: Jour. Exp. Zodl., vol. 1, no. 3, 1904. Woodruff: Biol. Bull., vol.
17, no. 4, 1909.
2L. L. Woodruff: Two thousand generations of Paramaecium. Archiv fir
Protistenkunde, Bd. 21, 3, 1911.
135
136 LORANDE LOSS WOODRUFF AND GEORGE ALFRED BAITSELL
not the crucial factor in the determination of the longevity of
paramaecia cultures. Further, and aside from this interesting
theoretical consideration, such a favorable ‘constant’ culture
medium would be valuable for breeding paramaecia in various
lines of experimental work, since it is clear from many investi-
gations that the reactions of paramaecia to various reagents, etc.,
are greatly modified by their past and present environment.®
In the present paper are briefly outlined the results which have
been secured, thus far, in an effort to answer the question sug-
gested above, and to provide a suitable culture medium which
investigators can employ in breeding cultures of this organism.
The animals used in this study were from the pedigree culture
of Paramaecium aurelia which one of us* has had under daily
observation for more than four years, and which has attained,
up to the present time (May 1, 1911), 2370 generations under
the conditions of a ‘varied environment,’ without conjugation
or artificial stimulation.
The favorable results secured first by Calkins® with strong solu-
tions of beef extract as a temporary stimulant for his degenerating
cultures of Paramaecium caudatum in infusions of hay, and later
by Woodruff* with Oxytricha fallax under similar conditions,
suggested the use of a weak extract of beef as a ‘constant’ cul-
ture medium. Further, beef extract should afford all the elements
required for the continued life of protoplasm. The results of
chemical analyses have shown that Liebig’s extract of beef is
3 For example, Greeley (Biol. Bull., vol. 6, 1904, p. 1), in a study of the effect
of various chemicals on the protoplasm of Paramaecium, wrote: ‘‘Maximal dilu-
tions can only be approximate, as the action of identical solutions is not the same
on paramaecia from different cultures, because no two are exactly alike in respect
to chemical composition and osmotic pressure;’’? and Miss Towle in a similar study
(Amer. Jour. Physiol., vol. 11, no. 2, 1904, p. 235) said: ‘‘The first step toward
a clearing of the haze that envelops the subject will be found, I believe, when an
effort is made to unify the conditions under which different investigators are
working.”
4 Woodruff: Loc. cit.
§ Calkins: (Archiv fir Entwick.-Mechan., Bd. 15, 1,1902). ‘‘The lean beef was
boiled in tap-water for fifteen minutes and allowed to stand until cool. The clear
fluid was then used without dilution.”
6 Woodruff: Jour. Exp. Zo6l., vol. 2, no. 4, 1905.
REPRODUCTION OF PARAMAECIUM IN BEEF EXTRACT 137
remarkably constant in composition, and therefore this standard
preparation, which is available for all investigators, was used
as the basis of our culture medium.
Having decided on Liebig’s extract of beef, it was necessary
to make a series of experiments to determine the strength of
solution which was most favorable for Paramaecium. The
solutions were made by weighing out one gram of the extract
and diluting this with varying amounts of distilled water. The
different concentrations of beef extract showed that a solution
of approximately 0.025 percent gave the best results. Accordingly
a quantity of this solution was made up which was sufficient to
provide culture medium for the organisms for a period of seven
months. This length of time was decided upon for this work
as the final results of Calkins’ experiments led him to conclude
that the cycle of Paramaecium caudatum, in a constant environ-
ment of hay infusion, was not of more than six months duration.?
The medium when made was put into over one hundred test
tubes, and these were plugged with cotton and sterilized. The
solution in the various tubes remained sterile until it was used,
and the inoculation of the medium with bacteria which were
transferred with the paramaccia afforded an ample supply of
food for the animals.
The regular experiment was begun on October 1, 1910, by the
isolation of a specimen from each of the four lines of the pedigree
culture I of Paramaecium aurelia at the 2012th generation.
Each of the organisms was placed on a depression slide in five
drops of the beef extract solution, and in this manner was started
a culture designated Paramaecium IB. This culture was con-
tinued by isolating an organism every day from each of the four
lines of the culture, and placing it in fresh culture medium on a
sterile depression slide. The number of divisions during the
previous twenty-four hours was recorded at the time of isolation.
From this record the average daily rate of division of the four lines
7It may be noted that animals from this same pedigree culture of P. aurelia
were bred on a ‘constant’ medium of hay infusion from February to June, 1909,
and died out after 107 days subjection to this condition. (Cf. Biol. Bull., vol.
17, no. 4, 1909, fig. 4).
138 LORANDE LOSS WOODRUFF AND GEORGE ALFRED BAITSELL
Ott Hov. Det. Jan. Feb. Mar Apr.
Fig. 1 Graph showing the rate of division of Paramaecium aurelia, Culture I
and Culture IB, from October 1, 1910, through April 29, 1911. The ordinates
represent the average daily rate of division of the four lines of the respective
cultures, again averaged for ten day periods. Culture I = ; Culture IB
of the culture, again averaged for five- and ten-day periods, was
computed, and the result is graphically shown in figs. 1 and 2.
The original pedigree culture on a varied environment was, of
course, continued, and served as a control for the culture on beef
extract, since the method of carrying on the two cultures was
identical, except that the medium used was not the same in each
case. The original pedigree culture was bred on infusions of
grass, hay, pond weeds, etc., made up with water from various
sources. The infusions were boiled before being used to prevent
the introduction of ‘wild’ paramaecia into the pure lines.®
The various preparations were kept in moist chambers to
prevent evaporation, and the temperature of the air in these
chambers was recorded by a maximum and minimum registering
thermometer. Obviously this method of recording the tempera-
ture gives only the extremes to which the cultures were subjected,
> For further details of the method employed, cf., Woodruff, loc. cit.
REPRODUCTION OF PARAMAECIUM IN BEEF EXTRACT 139
and does not take into account the length of time during which
any particular temperature was maintained. However, the
method is sufficiently exact for the problem at hand. Studies
on the relation of temperature to the ‘rhythms’ in the rate of
reproduction of paramaecia are now in progress. During the
first three months of the work, culture I (‘varied’ culture medium)
and culture IB (beef) were kept in different rooms, and therefore
during this time the cultures were subjected to different temper-
atures. From January 1 to the present time, both cultures were
kept in the same place and consequently each was subjected to
the same temperature.
There were, then, two pedigree cultures of Paramaecium, each
comprising four lines, being conducted simultaneously. One
of these had been bred on a ‘varied’ culture medium for forty-
one months, and was continued under the same conditions dur-
ing the following seven months, 7.e., to the present time. ‘The
other culture, isolated line by line from the first culture, was car-
ried on for seven months (to date) on a ‘constant environment’
of beef extract. The chemical composition of this medium was
identical from day to day as it was all made up and sterilized
at the same time. The only variation, therefore, in the medium
used for these organisms on beef extract was the fluctuations in
the bacterial flora due to infections from the air, and slight vari-
ations in the multiplication of the bacteria due to temperature
changes. This, however, was so small that it is negligible from
the standpoint of these experiments.
A study of figs. 1 and 2 gives a clear idea of the comparative
rate of division of the two cultures, and shows that, at the end of
the seven months work, the rate of division, and therefore presum-
ably the vitality, of the two series of animals is practically the
same. Neither of the cultures shows any indications of loss of
vigor, and the rate of division of each at the end of the experi-
ment is practically the same as at the beginning—such fluctua-
tions as have occurred being merely ‘rhythms’.
During the first three months of the work, the rate of division
of the beef series was considerably lower than that of the other
series, but this is obviously explained by the considerably higher
LORANDE LOSS WOODRUFF AND GEORGE ALFRED BAITSELL
140
aunyyng ! =] onyng «
IO} podvlaae urlese ‘saangpno aaryo
‘poqoofqns a1aM Samnjpnd at
Jdy Ie
{OJ9Q WOA] ‘Gy aIngyny pue JT ainypny
ad sutpuodsas109
YoIyM OF vinjetodur
424
ayy 10
t 9G} jo soul] mmoj oq} jo UOISTAIpP JO 94¥1
9} 99R.
‘uer
J (q,.) 2npRtedure, ssvi9oAv ay} OSTR puv
ASLIOAB OY} JUK
INAV VY} (@BAOGV) oOs[e pue ‘TIGL ‘6% [Ady |
‘OI6T
GSI
a!
REPRODUCTION OF PARAMAECIUM IN BEEF EXTRACT 141
temperature to which the latter was subjected during this period
(cf. temperature curves in fig. 2). During this period of three
months, culture I advanced from the 2012th to the 2188th gen-
eration, or 176 generations, whereas culture IB advanced from
the 2012th to the 2120th generation, or 108 generations. A
comparison of the rate of division from January 1 to April 29,
1911, when both cultures were subjected to identical temperature
conditions, shows that culture I advanced from the 2188th to the
2365th generation, or 177 generations, while culture IB advanced
from the 2120th to the 2287th generation, or 167 generations.
Therefore the net vatiation in the number of generations attained
by the two cultures during the last four months, when under the
same conditions of temperature, was only ten. Further, the ap-
pearance and behavior of the paramaecia in the two cultures
were identical.
It is evident, then, that the ‘constant’ medium of beef extract
employed has proved (during the seven months of this experi-
ment) to be practically as favorable a medium for the reproduc-
tion of this pedigree culture of Paramaecium aurelia as the ‘varied
environment’ medium,* and therefore, the conclusion seems jus-
tified that this culture of Paramaecium can, in all probability,
be continued indefinitely on this ‘constant’ medium. It there-
fore appears fair to conclude that it is the ‘composition’ of the
medium rather than the ‘changes’ in the medium which is con-
ducive to the unlimited development of this culture without con-
jugation or artificial stimulation.
It is not suggested that every culture of Paramaecium would
have the potential to attain more than two thousand three hun-
dred generations under the conditions of a ‘varied environment,’
nor is it suggested that every culture of Paramaecium would
thrive for over seven months on a ‘constant’ environment of beef
extract. ‘‘For undoubtedly there are strong and weak strains
’ Experiments are now in progress to determine if this culture of Paramaecium
will develop indefinitely on infusions of hay which are not made up the same from
day to day, 7.e., if hay infusion, in which there is a slight daily variation, may
not be substituted for the decidedly varied culture medium which has been used
during the past four years. .
142 LORANDE LOSS WOODRUFF AND GEORGE ALFRED BAITSELL
among Infusoria as among other classes of animals. Again, it
is possible that the different races of paramaecia which Jennings
has been able to isolate may have a physiological as well as a
morphological basis of distinction.’'° But it is believed that these
experiments clearly show that beef extract (in the concentration
used) is a suitable environment for the continued reproduction
of this pedigree culture of Paramaecium, and that beef extract,
in this or closely similar solutions, will prove to be a favorable
medium for use in many investigations on the physiology of
Paramaecium.!!
10 Woodruff: Biol. Bull., vol. 17, no. 4, September, 1909, p. 303.
1 After the completion of the seven months experiment (April 29, 1911), a new
lot of the beef extract medium was made up exactly the same as before and the
culture has been continued on this without noticeable change in its reproduction to
date of correcting proof, June 20, 1911.
MIGRATION OF RETINAL PIGMENT IN THE EYES
OF BRANCHIPUS GELIDUS
RUTH B. HOWLAND
FOUR FIGURES
INTRODUCTION
The general subject of pigment migration has received much
attention during recent years. While it is not the purpose of
this paper to deal with the movements of melanophore pigments,
we can not disregard the effect which the study of this problem
has had upon the study of the migration of retinal pigment.
There can be no doubt that an intimate relation exists between
the movements of the two forms of pigment, as there is no doubt
concerning the morphological similarity existing between pigment
spots and the eye spots of many of the lower animals.
The striking changes in coloration exhibited in chameleons
early led to investigations as to the physical factors involved.
Conclusive evidence showed that the movements of melano-
phore pigment was brought about largely by variations in light
intensity and temperature. The work of Carlton (’03), Parker
and Staratt (04), and Parker (’06) is sufficiently well known to
need only brief mention. Carlton found, in the case of Anolis,
the outward migration of pigment in the pigment spots to be
directly dependent on the action of the sympathetic nervous
system, and the inward migration simply a return to the resting
state.
In chameleons the migration was found to be under the control
of the spinal nerves.
As these observations have been made only upon the marine
crustacea, it was suggested to me by Professor C. W. Hargitt
that a series of comparative studies upon a fresh water crustacean
might be of some value. The occurrence of the Phyllopod,
Branchipus, in a pond near Syracuse, and its especially promi-
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. LI, No. 2
AuGust, 1911
143
144 RUTH B. HOWLAND
nent eyes made it a very favorable object for the investigation.
In the spring of 1908 these crustacea could not be found at Syra-
cuse, but material for experiments was obtained from Jordan,
a small town seventeen miles west, and this supply was supple-
mented by material from Potsdam and the West, and fresh speci-
mens from Syracuse in the spring of 1909. Experiments were
made and subsequent study conducted with a view to determining
the effect of varied heat and light on the migration of retinal
pigment in these eyes.
HISTOLOGICAL METHODS
The methods of preparation which gave the best results are,
in brief, as follows:
1. For killing and fixing, hot picro-acetie acid was used. The
animals were dipped in and held for a few seconds, then trans-
ferred to 80 per cent alcohol. Formalin proved unsatisfactory
for fixation, and also rendered the material hard to stain in some
eases. Corrosive sublimate fixation was also inferior to picro-
acetic.
2. a. Ehrlich’s haematoxylin in toto or section gave the best
results of any of the stains tried. If used for staining in toto, a
a period of at least sixteen hours was necessary. When the sec-
tions were stained upon the slide, a minimum staining of ten
minutes was allowed. The one stain was sufficient to show all
features clearly, but eosin was occasionally used as a counter
stain. .
b. Iron-haematoxylin with counter stain of Bordeaux red also
gave satisfactory results. The slides were treated as follows: left
in iron-haematoxylin, a minimum of fifteen minutes; plunged
into a FeCl, solution until sufficiently differentiated; washed in
water; stained in aqueous solution of Bordeaux red for three min-
utes; dehydrated; cleared and mounted.
c. Borax carmine failed to give the desired differentiation.
d. Haidenhain’s triple stain was temporarily satisfactory, but
was not permanent.
3. Xylol proved to be the best clearing agent. Cedar oil
was not so good. Material 7m loto required an hour at the mini-
EYES OF BRANCHIPUS GELIDUS 145
mum for clearing. Infiltration of at least one and one-half
hours was necessary. Paraffin melting at 62° C. permitted sec-
tioning at from 3 to 5 microns.
4. To see clearly the nuclei of the retinular cells which in the
normal eye are obscured by large masses of pigment, the follow-
ing method of depigmentation was used:
Two or three drops of hydrochloric acid were poured over a few
crystals of potassium chlorate. When the green color of the
evolving chlorine appeared, a few cubic centimeters of 70 per
cent aleohol were added. The eyes were allowed to stand in this
bleaching fluid over night and subsequent staining revealed the
retinular nuclei clearly.
NORMAL STRUCTURE OF THE EYE
Before entering upon a discussion of the experimental work,
a brief description of the normal eye is essential.
The eye of this form is one of the simplest of compound crus-
tacean eyes. It is stalked, and of such large size as to be a very
prominent feature. The structural elements are fundamentally
the same as those of other crustacean eyes, the unit being the
ommatidium.
The histology of the eye of Branchipus vernalis is described
in brief by Parker (’91) and others, and a more detailed account
of the eye of Branchipus stagnalis is given by Nowikoff (’05).
The outer surface of the eye is covered with a structureless
cuticula, presumably a secretion of the underlying hypodermal
cells. In Branchipus vernalis it is obviously facetted, having
the form of concavo-convex elevations.
According to Patten, the hypodermal cells in B. grubei are
of indefinite arrangement, but Claus and Nowikoff find in B.
stagnalis a regular arrangement which the latter shows diagram-
matically in the following way: the hypodermis covers the distal
end of each ommatidium as a cuticular cap made up of six equal
cells extending radially from the center to the circumference of
the circle. The nuclei are oblate, and lie upon one side. The
intervening space is filled by two cells with round nuclei. These
nuclei are extremely large and extend at an acute angle into that
146 RUTH B. HOWLAND
part of the cell lying between the cone cells. This arrangement
as well as the facetted condition of the cuticula (both disputed
points) have been observed in my preparations of the species
found here.
The cone cells are four in number, of equal size and regular
form. The crystalline cones are ellipsoidal and their mass in
comparison with that of the cone cells is large. The nuclei of
the cone cells were not found by Claus in the adult, and only
indistinctly seen in the larvae. Nowikoff, however, states that
they are well defined in his preparations, lie close under the erys-
talline cones at the promixal ends and are sphere shaped, while
Patten finds these nuclei lying over the distal end of the cells
like a cap.
In regard to the exact location of these nuclei, it seems to me
that there has been a great deal of confusion arising from incom-
plete knowledge of the histology of the hypodermal cells. Patten,
as I have previously stated, regarded the hypodermis as consist-
ing of a “layer of indefinitely arranged cells’ and the four cone
cell nuclei as forming a cap over the crystalline cones. These
nuclei, described as forming a cap over the crystalline cones,
might be those of the hypodermal cells in their characteristic
radial arrangement as shown by Nowikoff. If this were the case,
Patten’s description would agree with B. stagnalis as to the loca-
tion, but not as to the number of the hypodermal nuclei. Now-
ikoff, however, goes a step further in identifying the cone cell
nuclei.
I regret to have to say that in my own preparations I have been
unable to confirm Nowikoff’s or Patten’s statement as to the
location of these nuclei, though I have directed special attention
to this point. It is a most peculiar circumstance that cells of
this size and importance should fail to show clearly a definite
nuclear structure.
A peculiar condition in one series may be worth considering
in connection with this point. This series, cut at 5 microns,
shows, under the one-sixth inch objective, structures which even
upon critical examination appear as nuclei of the cone cells at
the level of the distal ends of the retinular cells. A slight enlarge-
EYES OF BRANCHIPUS GELIDUS 147
ment of the rhabdom at this place furthers the impression that
we are looking at nuclear structures. Similar structures, though
less evident, can be seen scattered elsewhere at various levels.
They are of comparatively large size and appear round and very
similar to the nuclei pictured by Nowikoff in B. stagnalis. Nu-
clear structures, if present in this level, would easily escape even
careful observation; for unless the section be extraordinarily
thin and exactly in the plane parallel to the rhabdom, the ends of
the pigmented retinular cells would obscure them.
However, a closer study under a one-twelfth inch oil immersion
lens shows these structures rather as reticulated fibrils of the
cytoplasm of the cone cells with scattered pigment granules
forming pseudo-nucleoli. Had these structures been constant
in all of my slides, I would have been assured of the fact that they
were nuclei, but their searcity, the reticulated structure of the
cytoplasm of these cells, and their appearance under the oil im-
mersion lens forces the conclusion that they are artifacts.
The rhabdom is slender and in continuation with the cone cells
(Nowikoff) ; it extends approximately three-fourths of the entire
length of the ommatidium terminated proximally by the base-
ment membrane, and is surrounded its entire length by the five
retinular cells. The retinular cells contain a large quantity of
pigment granules. Their cytoplasm is fibrillar, extending below
the basement membrane as fibrils. The nuclei are oval and lie
in the larger distal ends of the cells.
The nucleated basement membrane separates the retinal and
nerve-fiber regions. It is perforated where the retinal cells pene-
trate it (Nowikoff, ’05).
Blood corpuscles are found in varying numbers between the
ommatidia, and especially upon the ventral side of the eyes.
This was particularly noticeable in a large number of prepara-
tions, and is doubtless due to the close proximity of a large ventral
sinus.
A comparison of sections of the eyes of B. vernalis and B.
gelidus with the plates and descriptions of B. stagnalis given by
Nowikoff show no differences of structure. The nuclei of the
cone cells I was, of course, unable to compare.
148 RUTH B. HOWLAND
An article by Dietrich (09) came to my notice after the com-
pletion of the above discussion. It comprises an extensive
description of the facetted eyes of diptera, and contains points of
general interest in their bearing upon the problem under dis-
cussion. Mention is made of the fact that deep-sea crustacea
have a greater number of elements in each ommatidium—pos-
sessing as a rule seven—than do the crustacea of a more pelagic
habit, whose ommatidial elements have been reduced to five.
An explanation of this is offered by the fact that the difference
in intensity of light which these two extremes receive has possibly
led to the reduction as an adaptive feature.
Another feature which calls forth comparison with the structure
of the phyllopod eye is the location of the nuclei of the cone cells.
In some species of diptera, these are found directly beneath the
pseudocones, while in others they le further proximally. Their
appearance in the latter case closely resembles the structures in
my preparations already described, which closer study revealed
as artifacts.
EXPERIMENTS
An account of the experiments tried may perhaps best be
given in tabulated form:
SPECIES HOW KILLED LIGHT OR DARK TIME TEMP.
Deg. C
Bip weliduse av3a2 4: Corrosive sublimate Dark 9 hrs. 525
B. gelidus.. . : Corrosive sublimate Dark 8hrs. | 20.0
3. gelidus. ...... Hot picro-acetic Dark 43> hrs. 24.0
B. gelidus ... Corrosive sublimate — Diffuse light S8hrs. 17.0
B. gelidus .. Corrosive sublimate Diffuse light 9hrs. | 55.0
B. gelidus : Hot picro-acetic Sunlight Killed at 19.0
six o'clock
B. gelidus ae Hot picro-acetic Sunlight 12hrs. } 21.0
B. gelidus. . . Hot picro-acetic Dark 16hrs. | 21.0
B. vernalis Hot formalin Dark 5hrs. i
B. vernalis Picro-acetic Dark 5hrs. 2?
B. gelidus. . Hot picro-acetic Dark 30 min. 19.0
B. gelidus..... Hot picro-acetic Dark 15-20 min. | 19.0
B. gelidus........... Hot ipero-acetic Dark 2 hrs. 5.0
B. gelidus...........| Hot picro-acetic Dark 2hrs. | 21.0
EYES OF BRANCHIPUS GELIDUS 149
The animals were taken from the pond and experiments made
upon them without delay, for the character of all reactions was
noticeably modified by keeping them for twenty-four hours in
an aquarium. A loss of vitality, fading out of color, etc., became
evident within a day unless the aquarium was kept at very low
temperature, with a supply of soft mud in the bottom.
In all the experiments, the largest, most active specimens of
both sexes were chosen. The experiments in light and dark were
conducted as follows:
a. A few animals were placed in the direct sunlight, after allow-
ing them to stand in a warm room until the water had been grad-
ually raised to room temperature. In the early part of the
season this method was found to be impossible. The maximum
temperature at which they will live changes with the advancing
season, and while early in the season the room temperature of
21° C. will kill them almost immediately, later the maximum is
raised to from 26° C. to 29° C. The range of temperature be-
tween the optimum and the maximum for these animals, however,
does not seem to vary. With a higher optimum, a higher maxi-
mum temperature ensues, but the range between these two ap-
pears to have a certain constancy of 5-8° C.
b. On a cloudy day, an aquarium with a few specimens was set
out of doors and allowed to stand for nine hours. The average
temperature for that time was 5.5° C. and the few degrees of
variation above or below were not considered of sufficient value
to appreciably modify the result. The same experiment was
twice repeated with aquaria upon a table in the center of the room,
out of the direct rays of the sun, at temperatures of 17° C. and
20° C. for periods of eight hours.
c. In conducting the experiments in the dark, the greatest care
was taken to exclude every means of access of light rays. The
animals were placed in a tall glass dish, which was wrapped in
light-proof paper. This was then either placed in a tin box and
covered tightly, or left in the photographic dark room for the
required length of time. The paper was then removed, and a
flash of light thrown into the jar to assure the fact that all were
150 RUTH B. HOWLAND
alive. The water was at once drained off and the specimens
thrown into the killing fluid, heated to 60° C.
In preparing the eyes for sectioning it was found that the lia-
bility to injury from handling was greatly reduced by dehydrat-
ing and clearing the entire animal before separating the eyes from
the body. After infiltration was complete, the eyes were removed
from the body and imbedded with special attention as to orienta-
tion for transverse or longitudinal sections as desired.
CONDITIONS FOUND
A discussion of the conditions found in the material sectioned
falls naturally under three heads: first, the effect of ight; second,
the effect of heat variation; and third, the relation of pigment
migration to phototropism.
1. In these examinations care was taken to compare those
sections cut at equal thickness, and to select those parts of the
sections (long) where the plane of cutting was parallel to the long
axis of the rhabdom.
a. In eyes which have been killed after exposure to the sun for
a period of one and three-fourths hours the pigment is in the fol-
lowing condition: In the distal parts of the retinular cells the
granules are accumulated in great quantities, so that it is impos-
sible to distinguish the nuclei of these cells, which are located here.
The entire length of the rhabdom is closely protected with a
heavy layer of pigment, and here the lateral portions of the cells
are much clearer as compared with the distal part. In the prox-
imal third of the cells there are proportionately smaller quanti-
ties. Pigment is also heavily deposited below the basement
membrane, and for a short distance along the nerve fibers. In
eyes exposed to sun for longer periods, the conditions are identical.
(Figs. 1 and 3.)
b. In diffuse light for eight hours, the pigment granules are
still heavily deposited in the distal ends of the cells, and along
the rhabdom. ‘The lateral portions of the retinular cells, as in
the case of sunlight, are much clearer than the densely pigmented
region close to the rhabdom. The direct sunlight thus seems to
produce no appreciably greater response of pigment than the
diffuse light of a partly cloudy day.
EYES OF BRANCHIPUS GELIDUS 151
Fig. 1 Pigment conditions after exposure to sun until 6 p.m. on a sunny day
Rhabdoms covered with pigment.
Fig.2 Pigment conditions after remaining in dark four and one-half hour
Readjustment of pigment through a lateral migration leaves rhabdoms exposed
Fig.3 Transverse section of ommatidia after exposure to sun for five hours
Conditions as 1n fig. 1 (Oil immersion, camera lucida
Fie. 4 Transverse sec.ion of ommatidia after remaining in dark four and one-
half hours. Conditions as in fig. 2 One-twelfth oil immersion, camera lucida
152 RUTH B. HOWLAND
c. Eyes sectioned after remaining in the dark fifteen minutes,
twenty minutes, and a half hour, show no appreciable movement
or migration of pigment granules. The conditions do not vary
from those in diffuse light to any appreciable degree. Pigment
is heavily deposited for a short distance down the nerve fibers
below the basement membrane.
d. In the eyes which have been in the dark for two hours, the
pigment has begun its lateral migration and forms an intermediate
stage between the two extremes a and e.
e. The eyes which have been in the dark four and a half hours,
show the pigment closely packed in the distal ends. The granules
are not packed close to the rhabdom as in sun and diffuse light,
but are scattered laterally in the cytoplasm by a lateral migration
or a readjustment. The outline of the rhabdoms is thus not so
clearly defined by a heavy boundary of pigment granules, but
their surfaces are more exposed. ‘The entrance of light rays would
result in a more intense stimulus than under the previously
described conditions. A dense accumulation of pigment occurs
below the basement membrane and proximally down the nerve
fibers (fig. 2 and fig. 4).
f. In eyes left in the dark from eight to sixteen hours, these
variations in pigment location become less and less evident, and
in general the eye assumes the appearance of those killed in direct
sunlight. The distal pigment is closely packed around the rhab-
dom and the retinular nuclei, the pigment in the proximal part
lying close to the rhabdom. Eyes which have been in the dark
for eight hours or more were thus not considered in drawing con-
clusions as to the effect of light and dark.
The adaptive movement of pigment granules in these eyes would
thus appear rather as a rearrangement or readjustment of the
pigment granules in the retinular cells than as a pronounced
migration proximally in the dark, and distally in the light. A
proximal migration in the dark would result necessarily in an
accumulation of pigment below the basement membrane, which
would increase gradually from the light conditions to the com-
pletion of the migration. The pigment masses beneath the
membrane would thus vary in a graded series, a condition which
EYES OF BRANCHIPUS GELIDUS 153
does not oecur in my preparations. The pigment deposited be-
neath the membrane is quite as dense in sunlight as in dark.
The distal pigment in all my preparations is densely packed around
the rhabdom and the retinular nuclei, and shows no sign of move-
ment due to light or temperature changes.
The protective function of the pigment is accomplished in a
different way. A lateral movement of the granules exposes the
rhabdom in the dark, while an opposite movement in the light
brings them close against the rhabdom. Since these eyes have
no accessory pigment cells, which in other crustacean eyes serve
as a reflecting apparatus in the dark, the cytoplasm of the retin-
ular cells must perform this function.
So, unlike the eyes of Cambarus and Gammarus, the proximal
and distal migrations of pigment granules are not found in B.
gelidus, but are replaced by a lateral migration, outward in the
dark, centrally toward the rhabdom in the light, while the pro-
tective function of the pigment remains the same.
2. The temperature changes, so far as I have been able to note,
produced no appreciable effect on the migration of the pigment.
High temperatures were of course impracticable in experimental
work, causing almost immediate death. In eyes exposed to dif-
fuse light at 5.5° C. and 17° C. no difference could be noted in the
distribution of the pigment.
3. Experiments were made with a view to determine the de-
gree of phototactic response, and also its possible bearing on the
movement of the pigment.
a. A dozen animals, which had just been brought in from the
pond, were put into an aquarium in a large amount of pond water
at 21°C. Of this number, five were young specimens, the others
adults. Light from a 16-candle power lamp was thrown into the
aquarium from above with the following results:
11 were strongly positive. 1 female was indifferent.
The two youngest specimens made frantic efforts to approach
nearer the light beating their heads against the sides of the aqua-
rium. The light was then placed below the jar. Nine responded
at once by following the light. One male, one female, and one
young specimen were indifferent for a period of three minutes,
154 RUTH B. HOWLAND
when they too went to the bottom. If the light were held below
the jar, the tendency to orient themselves with their long axes
perpendicular to the source of light was stronger than the habit
of swimming with the appendages upward, and they stood upon
their heads making futile efforts to reach the light.
b. A larger number of animals was then put into the jar and
the same experiments repeated. The younger ones showed a
much quicker and more positive response than the adults. The
depth of the water effected no change in the character of the re-
sponse. Special attention was given to the occurrence of the
animals in their natural habitat, to determine whether the re-
sponses given in the laboratory were merely artificial or whether,
even in the natural environment, the direction of the light rays
had a noticeable effect on the location of these crustacea. On a
bright, sunny afternoon, when the water was 11° C. a much
larger number of specimens was found swimming on the west
side of the pond than on the east side. They were not, however,
very active, on account of the low temperature, but for the most
part were swimming slowly along the bottom about an inch above
the mud. For some time, the direction of the wind was thought
to have a marked influence on the distribution in the pond, but
the wind on this day was blowing from the northwest, and the
animals must have gone almost directly against the wind in order
to appear in such large numbers on the western side of the pond.
c. The temperature of the water in one jar was gradually raised
by the addition of small amounts of hot water. Pond water
was heated for this purpose to guard against the addition of any
chemical which might influence the response. When the water
reached 24.9° C. to 25.5° C., the light was placed above the jar.
The animals responded positively as in b. They became very
active, and the branchipeds greatly increased their rate of motion.
As the temperature reached 27° C., convulsive jerkings became
frequent, jerking the animals almost out of the water when they
came to the top in response to light. Between 27° C. and 29°
C. they became less and less active and sank to the bottom of
the jar. At 31° C. they were all dead. The increase in temper-
ature does not, therefore, cause a negative response, but merely
EYES OF BRANCHIPUS GELIDUS 155
a greater activity which finally leads to spasmodic muscular con-
traction and death.
Numerous reports of photomechanical changes in retinal pig-
ment have appeared referring to amphibia, cephalopoda, and
arthropoda. Exner (’91) and Parker (’97 and 799) record the
effect of changes in light intensity on retinal pigment in Palaemon
and Gammarus.
These reports have led without exception to the conclusion
that the same laws of migration which hold true in the case of
melanophore pigment apply as well to the migration of retinal
pigment under various degrees of illumination. The uniformity
of the effect of heat, however, has only recently been acknowl-
edged. Kiihne (’79) observed that the retinal pigment in frogs’
eyes in darkness was withdrawn further proximally in a high than
inalow temperature. Herzog (’05) confirmed and amplified these
results, and further stated that while increased heat produced a
proximal migration, decreased temperature a distal migration
between 0° and 18° C., above this temperature exactly the reverse
took place. The question of temperature influence was resumed
several years later by Congdon (’07) in the case of Arthropod
eyes. Experiments on Palaemonetes and Cambarus confirmed
the previously published results of Ktihne and Herzog. Response
to raised or lowered temperature in these crustacea is much weaker
than photomechanical response, and probably of no functional
importance, as the migration due to temperature change occurred
much above normal conditions.
As to the physiological importance of retinal migrations due
to varied light conditions, there is no doubt that the distribution
of the pigment along the sensitive parts of the eye protects it
from too intense illumination by the absorption of light rays;
while on the other hand, the withdrawal of pigment from them
gives easy access to the non-injurious rays of diffuse light. In
eyes which possess, in addition to the dark pigment cells, whitish
accessory pigment cells (Palaemon, according to Congdon, ’07)
these cells serve to reflect light rays into the rhabdom on with-
drawal of the pigment in diminished light.
156 RUTH B. HOWLAND
In addition to this protective result the migration of retinal
pigment has a bearing upon the phototropic response of animals
in which this occurs. Gammarus annulatus shows (Smith, ’05)
pronounced migration of pigment, and accompanying this a
marked change in behavior toward the light stimuli. The close
correlation of these two facts in matter of time, points towards
the fact that the change from negative phototropism to a marked
positive response is due to the exposure or protection of the rhab-
dom by movement of the pigment.
d. The effect of a moving light was then tried. A light was
held above the jar until the usual response was given, and the
number of negative or indifferent specimens counted. Then it
was moved slowly around the jar. In every case the positive
response was greater, and more definite with the moving than with
the stationary light.
e. A larger number of animals was placed in an aquarium and
allowed to stand in the dark for an hour at room temperature.
They were then exposed to light. They remained indifferent for
a short period before any definite response was noted, but then
gave a positive response.
Another aquarium was tightly wrapped in black paper and
made thoroughly light-proof. This was placed out of doors
(about 19° C. and allowed to remain for five hours. It was then
taken to a photographie dark room and opened under a light of
16-candle power intensity. The majority of the specimens were
negative for a period of from two to three minutes, but after
remaining in the light, gradually became positive as under the
usual conditions.
The same specimens were at once placed in the dark and left
for two hours longer. When exposed to the light, the negative
reaction was even more definite than before.!
‘In a recent paper on phototropic responses of Branchipus (McGinnis, ’11)
the statement is made that Branchipus is never negative to light, even after ex-
posure to darkness, but no record is made of the time necessary to bring about
the normal response after this exposure. The peculiar ‘‘reversal of geotropic
response’’ recorded in the same paper (p. 237) may have been influenced by nega-
tive response to light, the other positive responses to gravity being, as suggested,
simply the natural falling of bodies.
EYES OF BRANCHIPUS GELIDUS 157
The reversal of the response in my experiments is obviously
related to the movement of the pigment in the eyes. As the
previous results have shown the condition of the eye at the end
of five hours to be one in which the sensitive rods are exposed to
the fullest extent, the sudden entrance of a strong light would
naturally cause a negative reaction. The fact that in the first
experiment an exposure to 16-candle power light did not cause
this negative response is doubtless due to the incomplete migra-
tion of the pigment in the shorter period of time.
SUMMARY OF RESULTS
1. The effect of light and dark on the movement of pigment
granules in the eye of Branchipus gelidus is in the nature of a
readjustment rather than a proximal and distal migration.
2. The distal pigment is not influenced by variation in light
intensity.
3. In light, the pigment granules collect closely around the rhab-
doms, protecting them from too intense stimulation.
4. In the dark, the granules move laterally and are readjusted so
that they become more evenly distributed through the eytoplasm
of the retinular ¢ells.
5. The time occupied in complete readjustment is between four
and one-half and five hours.
6. The cytoplasm of the retinular cells serves as a reflecting
apparatus in a weak light in the absence of accessory cells.
7. Changing temperatures have no appreciable effect upon pig-
ment migrations, higher temperatures causing almost instant
death.
8. Branchipus gelidus is positively phototropic. Animals ex-
posed to light after remaining in the dark five hours were nega-
tively phototropiec.
158 RUTH B. HOWLAND
BIBLIOGRAPHY
Cariton, F. C. 1903 The color changes in the skin of the so-called Florida
chameleon, Anolis Carolinensis Cuv. Proce. Amer. Acad. Arts and Sci.,
vol. 39, no. 10, pp. 237-276, 1 pl.
DAHLGREN AND Kepner 1908 Principles of animal histology.
Dierricu, WitaeELM 1909 Die Facettenaugen der Dipteren. Zeit. fiir wiss.
Zool., Bd. 92, Heft 3.
Hay, O. P. ann W. P. 1889 A contribution to the knowledge of the genus Bran-
chipus. Am. Nat. February, vol. 23, no. 266, pp. 91-95.
Herzoc,H. 1905 Experimentelle untersuchungen zur Physiologie der Bewegung-
vorgiinge in der Netzhaut. Arch. f. Anat. u. Physiol., Physiol. Abt.
Jahrg. Heft 5-6, pp. 413-464, Taf. 5. (Original not available.)
Ktune, W. 1879 Chemis*he Vorgiinge in der Netzhaut. M. L. Hermann,
Handbuch der Physiol., Bd. 3, Theil 7, pp. 235-3842. (Original not
available.)
McGinnis, Mary O. 1911 Reactions of Branchipus serratus to light, heat and
gravity. Jour. Exper. Zoél., vol. 10, no. 2, pp. 227-239.
Nowrkorr, M. 1905 Uber die Augen und die Frontalorgane der Branchiopoden.
Zeit. fiir wiss. Zodl., Bd. 79, Heft. 3, pp. 432-464, Taf. 2.
Packarp, A.S. 1878 A monograph of the phyllopod crustacea of North America,
with remarks on the order phyllocarida. U.S. Geol. and Geog. Survey
of the territories of Wyoming and Idaho, 1879, pp. 295-514. Part 1,
39 pl.
Parker, G. H. 1891 Compound eyes in crustacea. Bull. Mus. Comp. Zodl.,
Harvard Coll., vol. 21, no. 2, pp. 45-140, 10 pl.
1897 Photomechanical changes in the retinal pigment cells of Palae-
monetes, and their relation to the central nervous system. Bull.
Mus. Comp. Zo6él., Harvard Coll., vol. 30, no. 6, pp. 273-300, 1 pl.
1899 The Photomechanical changes in the retinal pigment of Gam-
marus. Bull. Mus. Comp. Zo6l., Harvard Coll., vol. 35, no. 6, pp. 141-
148, 1 pl.
1906 The influence of light and heat on the movement of the melano-
phore pigment, especially in lizards. Jour. Exper. Zoél. Vol. 3, no.
3, pp. 401414.
Parker, G. H. and SvarattT,S. A. 1904 The effect of heat on the color changes
in the skin of Anolis Carolinensis Cuy. Proc. Amer. Acad. Arts and
Sci., vol. 40, no. 10, pp. 455-466.
Rogers, C. G. 1906 A chameleon-like change in Diemyctylus. Biol. Bull.,
vol. 10, no. 4, March.
SmitH, Grant 1905 Effects of pigment migrations on phototropism of Gam-
marus annulatus. Am. Jour. Physiol., vol. 13.
THE DEVELOPMENT AND FUNCTION OF
VOLUNTARY AND CARDIAC MUSCLE IN
EMBRYOS WITHOUT NERVES!
DAVENPORT HOOKER
From the Sheffield Biological Laboratory of Yale University
FIFTEEN FIGURES—ONE PLATE
INTRODUCTION
The influence of the nervous system upon development has
been much studied and discussed, and it is natural, owing to the
complicated nature of the subject and the close relation of the
nervous system to the rest of the body, that the views that have
been held have been diverse and conflicting.
The literature on this subject up to 1904 has already been fully
reviewed by Harrison (’04) and Goldstein (’04), and the reader
is referred to their papers.
Harrison (’03, ’04), using much earlier stages than Schaper
(98) had used, removed the entire spinal cord from embryos of
Rana sylvatica, virescens and palustris before visible differen-
tiation of the axial musculature had begun and found that the
muscle tissue subsequently differentiated normally. The irri-
tiability of the nerveless muscle was not fully tested, however,
though one of the operated embryos did give response to uni-
polar electric stimulation, showing its independent. irritability.
Harrison also found that muscle tissue would develop normally
when embryos of the frog were kept anaesthetized in acetone-
1The results detailed in this paper were reported in brief at the May meeting of
the Society for Experimental Biology and Medicine, 1910, and an abstract appears
in its proceedings. A more detailed account was reported before the American
Association of Anatomists, at the Ithaca meeting in 1910.
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, No. 2
159
160 DAVENPORT HOOKER
chloroform during their development. In this way, the motor
nerve impulses normally transmitted to the developing muscle
were eliminated.
Goldstein (04) published a comprehensive paper after repeat-
ing Shaper’s experiments with improved technique. The embryos
which he used were of the same age as Shaper’s and, even though
the whole spinal cord and hind-brain were removed, the experi-
ments are therefore open to the same objection as Shaper’s, viz.:
that the first nervous connections had been established before
the operation. Goldstein found that the operated embryos de-
veloped normally and that their muscle tissue would respond
to direct stimulation.
Wintrebert (’03—’05) removed the cord from frog and Ambly-
stoma embryos at various stages and found that they developed
normally. He described the presence of contractility in the muscle
tissue before the nerves reach it. He states that he observed
not only contraction of the myotomes, but also movement on
direct stimulation of the fingers of the hind extremity of the frog
larva, from which extremity all nerve tissue had been removed.
The most recent work directly bearing on this problem is that
of Paton in 1907. He concluded that the first movements of
the embryo take place before there is any true nervous connection
between the muscle tissue and the central nervous system, and
used this as an argument in favor of the Hensen theory of nerve
development.
The purpose of the experiments detailed in the present paper
was to determine the following: (1) Does differentiation of muscle
fibrillae and the establishment of the nervous connection with
the central nervous system precede or follow the acquisition
of contractility in the myotomes in the normal embryo? (2)
Will voluntary muscle which has developed without the influence
of nerves respond to stimuli and, if so, must the stimuli be applied
directly or may they be transmitted by non-nervous paths?
(3) Will cardiac muscle differentiate and function independ-
ently of nervous control? (4) Will the gross form of the heart,
developing under such conditions, be normal?
MUSCLE IN EMBRYOS WITHOUT NERVES 161
METHODS
The operations were performed on frog embryos at the stage
of development immediately following the closure of the neural
folds (fig. 1). At this time they are from 2.25 to 3.75 mm.
in length, according to species, and there are absolutely no
nerve fibers in the central nervous system nor any traces of
peripheral nerves. The embryos operated upon by Shaper and
Goldstein were much older and the chances of nerve contamina-
tion were correspondingly greater.
The instruments used were: a pair of Noyes’ iridectomy scis-
sors, an iridectomy knife, forceps and needles, the points of all
being ground down to a more than hair-like fineness. All oper-
ations were performed under a Zeiss binocular microscope.
Fig. 1 Embryo of R. sylvatica, to show the stage of development used in
the beginning of the experiments. 93. (After Harrison.)
For the study of the voluntary muscle, it was necessary to
remove the source of nerve supply to the myotomes. The cord,
hind-brain and skin of the dorsal region were removed by Harri-
son’s method. The wound surface produced by this operation
is so small and heals so readily, that no skin graft was made.
For the experiments to study cardiac muscle, a much more
complete operation is necessary. Not only must the cord,
hind-brain and skin of the dorsal region be removed, but the fore-
and mid-brain, the primordia of the cranial ganglia (fig. 2) and
the skin of the entire head region must be taken out as well.
‘The cardiac plexuses receive such a large contribution from
the tenth nerve, according to the work of Kuntz (’09), that it
is especially necessary to extirpate its ganglia. In operating,
the first step is to remove the dorsal structures as in the previous
162 DAVENPORT HOOKER
experiment. The embryo is then laid on its side and the skin
cut from the dorsal wound surface to the ventral aspect of the body
behind the branchial region. When this has been done on both
sides, the skin is dissected away from the entire head down to
the suckers and removed by a circular cut just above them. The
remainder of the nervous system of the embryo which has not
been previously removed is thus completely exposed and may be
taken off with the knife or needles.
As may be well imagined, the cut surface in such an operation
is considerable, and, to prevent disintegration, a skin flap from
the abdomen of another embryo is grafted over {the wound.
Fig. 2 Diagrammatic sketch to show the cranial ganglia present at the stage
of operating—They are, from before backwards, the ganglia of the 5th, 7th—8th
complex, 9th and 10th.
The flap is held to the embryo by piling silver wire about it to
prevent movement, after the method of Born.
The embryos were operated in 0.5 per cent saline solution and
were kept in it for one or two days after. They were then trans-
ferred to water.
Histological examination showed that the nervous system was
entirely removed (fig. 3), except in four cases in which a small
portion of the infundibulum was found to have been left in. These
cases were carefully examined for nerves, but no trace of any
were found. In view of the fact that no differences were
found between the action of the heart in these and in the embryos
absolutely without any nerve tissue, they were not rejected.
MUSCLE IN EMBRYOS WITHOUT NERVES 163
Mechanical and electrical stimulation were used to determine
the irritability of the voluntary muscle. Theembryos were stimu-
lated with a human hair, according to Coghill’s method, and with
finely pointed steel needles. All stimulating was done under
the binocular microscope.
For observations upon the heart, the embryos were placed on
the stage of an ordinary microscope and the region just behind
Fig. 3 Diagrammatic cross-section of the body of an embryo of R. palustris
from which the entire nervous system was removed. H, Heart; N.C., Notochord;
notice the absence of the spinal cord above it.
the mouth watched for the beat. This method is a very accurate
one, as the slightest action of the heart is noticeable by the move-
ment of the skin. At later stages in palustris embryos, the
skin becomes transparent and the heart itself may be seen.
The embryos were killed in a sublimate-acetic mixture, sec-
tioned and stained either with iron haematoxylin or with Teld’s
molybdic acid haematoxylin. Those stained by the former
method were usually counterstained with Congo red.
164 DAVENPORT HOOKER
The experiments detailed in this paper were performed on the
embryos of Rana sylvatica, palustris and pipiens. The embryos
of Rana sylvatica are the least desirable for operating, owing to
the strong dorsal curvature of the body. Rana palustris and
Rana pipiens have nearly straight backs, but in the former there
is a greater contrast between the nervous tissue and the surround-
ings, so that they may be easily distinguished from one another
even macroscopically. For this reason palustris is preferable
to the more diffusely colored pipiens.
MOVEMENTS OF NORMAL EMBRYOS
The movements of frog embryos in response to stimulation
vary greatly according to their age. If embryos are carefully
watched in their development and stimulated at short intervals,
a stage will be found where some will respond to stimulation and
some will not. The reaction is characteristic. As pointed out
by Harrison (’04) “it consists of a sharp tonic contraction of the
myotomes on the same side of the body and immediately at the
point of the application of the needle prick.”” It is necessary that
the stimulus be given with a needle sharp enough to penetrate
to the myotomes. The stimulus, then, is a direct one, but it is
likely that the mechanical effect of the contraction of one myo-
tome may sometimes stimulate neighboring myotomes to contract.
This non-nervous type of response is very different from the
nervous type which is found in slightly later stages. The char-
acteristics of this later type were pointed out by Harrison in 1904
and have since been more thoroughly studied by Coghill (709).
At this stage, the response is asymmetrical, the contraction first
being away from the side stimulated, then toward it, the result
being a swimming motion.
A number of embryos taken at the first appearance of the early
or non-nervous type of response, some of which moved and some
did not, were fixed separately and cut into serial sections. On
examination, it was found that the development both of those
which responded and of those which did not was at essentially the
same stage. In both, the fibrillation of the voluntary muscle
MUSCLE IN EMBRYOS WITHOUT NERVES 165
had begun and motor nerve connections between the spinal cord
and myotomes were present. Thus it was evident that contrac-
tion on stimulation is a function normally acquired in the axial
musculature of the frog just after the beginning of the process
of fibrillation and after establishment of the connection with
the central nervous system. In the muscles of the limbs the
case is different, for Braus (05) found that the muscles in Bom-
binator do not respond to stimulation until a much later phase
of development is reached and not until spontaneous contraction
takes place.
Harrison (’04) showed that fibrillation was not dependent on
nervous connection. For proof that contraction is not depend-
ent on it, we must turn to the experiments.
MOVEMENTS OF CORDLESS EMBRYOS
Effect of mechanical stimulation
For the experiments on mechanical stimulation of voluntary
-muscle, 77 embryos were operated, of which 34 died, were abnor-
mal, or were imperfectly operated. Of the remaining 43 embryos
which showed no abnormalities, 27 moved on stimulation and 16
did not. It is interesting to note that three-quarters of the latter
died later from some organic deficiency, while less than half of those
that moved failed to live, indicating, perhaps, that the relatively
large number which did not react as compared with the number
which did was due to vital disorganization of some sort.
Each of these embryos was placed on the stage of the binocular
every twelve to twenty-four hours and stimulated with a human
hair (Coghill, 09) or with a very finely pointed needle. The
former method was finally rejected, since the operated embryos
did not respond to skin stimulation as do normal ones with an
intact reflex are. The latter method was very satisfactory, as it
permitted the direct stimulation of very minute areas of muscle
tissue without any profound injury to the embryo.
Of those embryos which responded to stimulation, the irri-
tability began on the first day after the operation in 9 individuals,
on the second in 16 and on the third in 2. In no ease did any
166 DAVENPORT HOOKER
embryo begin to show reactions later than the third day after
the operation. The reaction was a single quick contraction
toward the side stimulated with the point of stimulation as the
center of contraction. In no case was there a response unless
the needle point penetrated the skin. Three specimens gave a
tremor-like response, but microscopic examinations showed these
to have nerve tissue present in the trunk region. Spontaneous
movement was never observed in operated individuals. This is
contrary to the results of Shaper and Goldstein, both of these
authors having claimed that spontaneous movement is to be
observed in the ventral half of embryos which have been cut
longitudinally. I have repeated their experiments, operating
at an earlier stage, however, but have not been able to confirm
their results in this respect.
One of the most striking features of this experiment is that the
embryos do not continue to be irritable for an indefinitely long
period but cease to respond after a day or two. The irritability
to mechanical stimulation lasted but one day in 20 individuals,
two days in 5, and but one embryo responded during a three-
day period and one during a period of four days.
Embryos killed at the end of the first day and examined for
their histological structure, show that muscle fibrillation is well
under way and that the entire development of the body is at the
same stage as those of normal embryos in which contraction in
response to stimulation is beginning to occur. The nervous
system is, of course, lacking, but in general the stage of develop-
ment is the same.
Microscopical examination of the embryos killed and preserved
from six to eight days after the operation, shows, as brought out
by Harrison, that the muscle differentiation is practically com-
pleted, though not quite so fully as in the normal embryo, that
there is no nervous tissue posterior to the mid-brain and that,
with the exception of the absence of the hind-brain and cord, the
embryos are essentially normal. The principal abnormal feature
of these embryos is their oedematous condition, plainly visible
in the embryos as a whole and shown microscopically by the
presence of many vacuoles in the muscle tissue.
MUSCLE IN EMBRYOS WITHOUT NERVES 167
Effect of electrical stimulation
For the experiments on electrical stimulation of voluntary
muscle, 15 embryos were operated, all of which remained in the
best of condition until killed. They were each stimulated, at least
once daily, by a weak faradic current. The embryos were care-
fully watched through a binocular microscope while the stimu-
lation was applied. Each embryo was placed on a small plat-
inum plate in water, a fine platinum wire being used to form the
other electrode. On each stimulation, the embryo responded
by a single tonic contraction toward the side stimulated, the
point of stimulation being the center of contraction. The type
of contraction exhibited was exactly the same as that shown
when the operated embryos were stimulated mechanically.
The embryos were killed after four or five days, during which
time response to mechanical stimulation had ceased. Up to the
time they were killed, all the specimens responded in the manner
described above whenever stimulated electrically.
Microscopical examination of sections of these embryos shows
that the cord was entirely removed. The embryos were in every
respect similar to those used in the experiments on mechanical
stimulation.
EXPERIMENTS TO DETERMINE THE MODE OF TRANSMISSION
OF THE STIMULI
Wintrebert (04, ’05) has described a ‘sensibilité primitive’
existing in embryos of Rana esculenta during a period of four
days beginning at the stage following the closure of the neural
folds. He says that if an embryo during this stage is cut through
the back in the posterior part of the body, so that the spinal
cord, notocord and part of the yolk are severed, leaving the two
halves connected by the skin and yolk of the ventral region only,
the front part will respond on stimulation of the posterior part.
After the heart begins to beat, a cut near the anal region severs
the animal into two independent parts, of which the anterior
will not respond on stimulation of the posterior. Up to this time
168 DAVENPORT HOOKER
there are no peripheral nerves, and the yolk will not transmit
stimuli, so Wintrebert concludes that it must be the skin which
acts as the transmitter.
To test these results, three sets of experiments on frog larvae
and additional ones on Amblystoma were made. They were (1)
a repetition of Wintrebert’s experiments, (2) exposure and stimu-
lation of the yolk, and (3) girdling of the skin around the body
and stimulation of the posterior half with and without cutting
the cord.
Embryos with the cord cut. In repeating Wintrebert’s experi-
ments (table 1 and fig. 4), it was found that no reaction was ob-
tained in normal embryos of the first stage named by him. A
4.5 mm. Rana palustris embryo in which stage the tail bud is
quite pronounced was the smallest from which a response was ob-
tained. The discrepancy may, however, be due to the difference
between European and American forms. Furthermore, in all
TABLE 1
Repetition of Wintrebert’s experiments
EMBRYO SIZE RESPONSES
| mm | | | | |
ike 4.25 |0/0/0|0/0/0/0/0)0/0)0* 0
2a 4.50 |0/0/0/04 0/+70/0/0/0/0/0
Se ORM acs: 4.75 |0/0/0/0/0/0/0j;0/0/0]0]0
AE Dees. | 4.75 |0/0)/0/0/0/0/0/0/0/0)0}0
is a en arom ete | 5.00 }0/0]*/|*/0/+970/+970/0/0/0
Honbuan nearest: EU CO 0) Herel AO eaters) (2) | | CON er
sar eecn ee gat ee see] 6200) [ESE 94-14-59] 0) | 01/00 0/010
Se eee al) 16200 0/07 0/0/0)0/0)0/0)0/0/0
|
|
Intervals of stimulations 15 seconds.
*Spontaneous contraction without stimulation. Where this is found above a
negative response, it indicates that a contraction took ,place in the interval of
stimulation following.
+ Contraction slow in following stimulation.
°Contraction obviously due to shaking.
Total number of stimulations............ oe ees Ob
Total number of positive responses. 7 aS eer SLO
Total number of spontaneous Contractions, seta 5 5
Total number of contractions due to shaking.................. epee
MUSCLE IN EMBRYOS WITHOUT NERVES 169
but one case, the number of times when responses were obtained
from operated embryos was much less than the number of times
when no reaction occurred. The older embryos frequently move
spontaneously, however, and the probabilities are that some will
move during the time that they are under observation. Many
times the embryo moved just before it was to have been stimu-
lated. That the movement may occur apparently as a result of
stimulation, when it would have taken place without it, is abso-
lutely certain. Several times, while an embryo was being stim-
ulated at regular intervals and a large number of reactions was
being obtained, the omission of two or three stimulations showed
that the embryo was contracting more orlessrhythmically, whether
stimulated or not, the periods of the rhythm being the same as
the intervals of stimulation.
This, however, is not the only explanation which can be offered
for the contractions which take place. On careful analysis of
the conditions preceding the contraction, three factors can be
seen at work: (1) tension of the skin over the whole body pro-
duced by pressure with a dull needle, (2) shaking the embryo
during the stimulation, and (3) hindering the locomotion of the
embryo due to the action of its cilia. Whenever a needle sharp
enough to penetrate the skin without increasing the tension
elsewhere was used, no contraction resulted unless the myotomic
area was stimulated. The shaking of the operator’s hand will
clearly cause contractions, so that great care must be taken to
avoid this. Frequently the needle sticks after penetration and
the embryo is shaken as it is withdrawn. This also produces
contractions. Where the embryo’s progress, caused by its own
cilia, is suddenly stopped, a contraction almost always ensues.
This is avoidable by slowly checking the motion.
Stimulation of the yolk
In the second series of experiments, an opening was made in
the skin at the side of the animal (fig. 5) and the yolk stimulated
through it. Out of a total number of 100 stimulations, 29 con-
tractions were obtained. Of this number, 14 were obviously
170 DAVENPORT HOOKER
due to shaking the embryo by running the needle through the
body to the skin of the opposite side. This impaled the embryo,
withdrawal of the needle was diffieult and in every case the embryo
was shaken. One embryo was contracting freely between the
stimulations and the large number of responses given by it may
be accounted for to some extent in this manner. Table 2 demon-
strates the details.
The results show that the yolk, per se, as Wintrebert stated,
will not transmit impulses, but contractions may result from shak-
ing the embryo during stimulation.
6
Fig. 4 Diagram of a frog embryo to show position of cut (by dotted line) used
in the repetition of Wintrebert’s experiments.
Fig. 5 Diagram of a frog embryo to show opening through which the yolk was
stimulated in the second series of experiments to determine the mode of trans-
mission of stimuli.
Fig. 6 Diagram of a frog embryo to show “‘girdling”’ of the skin around the
body and the area in which stimuli were effective when the cord was uncut.
Girdled embryos
In the third series of experiments, the skin only was cut around
the entire body, thus girdling it (fig. 6).
Stimulation in the region of the cord (within the dotted area
indicated in the figure) when the latter is uncut, will nearly always
produce a response in the anterior end of the body (table3). If the
cord is cut, there is no transmission of stimuli from the posterior
part to the anterior. When the skin is girdled, the cord alone may
serve for the transmission of mechanical shocks to the anterior
MUSCLE IN EMBRYOS WITHOUT NERVES
half, such as those produced by shaking the embryo.
171
If the
cord is also cut, the yolk has not consistency enough to admit
of shaking the anterior half by disturbances in the posterior, con-
sequently no contractions result.
TABLE 2
Stimulation of the yolk only
EMBRYO! SIZE RESPONSES
gee 5.00 |+/+| 0/0 Pe Ae a] 04 0/0 +%+/0/ 0
gions 5.50/0 +40 /0|/+*0/0/0/0/0/0/0)/0\0 0/+*0/0/0/ 0
hone 5.75|0/0)0)0/0/0/0/0/0/0\0l+4+4 0 00 44/0/ 0/0
ae. 6.000 0 00/00/00 0 +1440 +40 /F} 0 +10 0/0
Ges 6.00/0/0/0/O}+*+4+ 544444444 0 00 +*/0|+*10] 0/0
Intervals between stimulations 15 seconds.
{Spontaneous contraction without stimulation.
*Needle pierced animal and embryo was shaken.
Embryo 1 contracted frequently between stimulations.
otalénumberkoras tins tlOMG esereceyteas i cetsrele ce ets tetedepete avatcteiey tose ie elet ciel 100
Total numberof positive TEsPONseS soci 150.
Fig. 9 Eudendrium. Three restitution masses interconnected, four days old;
two masses with coenosarcal outgrowths, es;7s, central part of original restitution
mass. Other lettering as before. 90.
Experiment August 1. Colony pressed out at 3:30 p.m. Tissue
in comparatively small amount is collected in watch glasses and
allowed to settle. Water is gently drawn off and fresh added
without disturbing the layer of ‘sediment’ that clings tothe bottom.
Glasses transferred to bowls after thirty minutes: transferred
again 6 p.m. On the following day all the larger masses dead or
dying. Many of the very small masses are also dead. Still
there are very many small masses, a fraction of a millimeter in
diameter, that are alive. These are more or less spheroidal and
covered with thin perisarc, some attached to bottom of glass,
304 H. V. WILSON
some to cover glasses that had been put in the watch glasses. On
August 3 many of these masses are still alive, surrounded by peri-
sare, but they have not thrown out coenosareal processes.
In this experiment a small amount of tissue was pressed out.
and until the details in the method of treatment are more precisely
marked out, this is certainly a safe step. It will be noticed that
the tissue was left 7m situ where it was first deposited on the bot-
tom of the glass. The results indicate that this is not a good
method for the larger masses. And yet it seems desirable for
the tissue to establish some connection with the bottom, and this
it will not do if disturbed and dislodged too much. The experi-
ment records show clearly what a great influenceapparently slight
differences in the treatment had on the vitality of the fusion
masses, and how much in the dark I remained as to what details
were good, and what bad.
Experiment August 2, a. Colony pressed out at 4:15 p.m.
Tissue collected in a watch glass, shaken gently to the center, and
water changed several times, each time the tissue being stirred up
considerably by the pipette current. Glass transferred to bowl 5
p.m.; transferred again 6 p.m. Tissue does not cling to the
bottom, as it does when left undisturbed where it first settled.
At 7. p.m. the tissue coheres sufficiently for pieces 2 to 3 mm. wide
to be sucked up with large pipette. Other smaller pieces about
1 mm. wide are sucked up. All pieces are thin, about 4 mm.
thick. These pieces, forty-five in number, are scattered over
the bottom of three bowls.
On August 3 at 9 a.m. the masses are alive and of good color.
Some are free, some slightly attached to the glass. The latter are
freed and all are gently transferred with large pipette to fresh
bowls of water. They resemble the Pennaria mass shown in fig.
14. Examination shows that the perisare has formed over parts
of many plates, but in other places while the surface of the plate
is smooth, no perisare can be seen. In still other places the con-
tour is rough, the periphery here consisting of rounded cells. Bac-
teria are present, here and there in swarms, but not much of the
tissue is dead.
BEHAVIOR OF DISSOCIATED CELLS 305
On August 7 a good deal of the tissue is now dead. But a
large number of the pieces include lobes and nodules of living
tissue surrounded by perisare. Two of the plate-like masses
show each a coenosarcal outgrowth of considerable length, about
like those of fig. 4. On August 8 two other masses show each a
similar outgrowth. On August 9 another mass has developed an
outgrowth, which soon becomes sickly, losing its well marked
layers and developing in the interior numerous dark masses. On
August 11 two other pieces show each a coenosarcal outgrowth.
These outgrowths are horizontal and creeping and each bears a
vertical branch. Both are sickly as is shown by the fact that the
layers are not everywhere uniformly differentiated, but in places
appear to be breaking up, while in other places the tissue is
densely concentrated.
The method practised in this experiment is evidently good and
yet too much of the tissue dies leaving the surviving masses slow
to transform. In the hope of stimulating these masses they
were given a liberal supply of pure oxygen on August 6, but with
no discoverable results.
While an effort was made in this and the other experiments to
pick out pieces large enough to be distinguished with the eye,
some pieces of gonophores and tentacle fragments remain in the
cultures. Some of these are incorporated by the plasmodial masses.
Many others undoubtedly die without being incorporated.
Another structure too deserves mention as being occasionally
present in the cultures. In cutting up the hydroid, many
hydranths are snipped off at the very base. Some of these get in
the cultures and escape notice. I have found that when such
hydranths are isolated a process of reduction takes place analo-
gous to that described by Schultz for hydra (’06), the hydranth
gradually becoming in the course of a few days a mouthless
spheroidal body. Apparently such a process goes on in the
pressed out cultures more rapidly, for oceasionally spheroidal
bodies are seen quite like the reduced hydranths just referred
to. I have moreover several times found such a body embraced
by a plasmodial mass. Where bodies like gonophores, tentacle
fragments, and reduced hydranths (or the bodies that look like
506 H V. WILSON
such) become surrounded by the plasmodial tissue, it is a question
what becomes of them. As already said, the formation of the
perisarc, [ am inclined to think, cuts out such bodies from the
comparatively homogeneous material round which it forms.
Pictures are sometimes had which indicate that possibly such
bodies are attacked by the plasmodial tissue, the latter invading
and absorbing them. Again a stray fragment of stem perisare
from the colony used may get into the cultures, and if some of the
coenosare has been left inside, it may form a regeneration knob
at one end of the piece. I have seen a piece or two of this kind.
Such a fragment might very well regenerate a hydranth in the
midst of the plasmodial masses. But fragments of this sort are
easily distinguished from the plasmodial masses or nodules of
the latter.
Experiment August 2, b. A colony was pressed out and the
tissue allowed to settle on a cover glass immersed in a watch glass.
In transferring, the cover glass was always kept in the watch
glass, and thus the tissue was not directly exposed to the air.
The material settling on the cover soon transformed itself into a
multitude of small, more or less spheroidal masses, a fraction of ¢
millimeter in diameter. They went so far as to form a perisare.
A large number of them died about one day after preparation,
but many remained alive for days; were still alive on August 9.
They had not sent out coenosareal outgrowths, but as was learned
later from sections the originally solid mass in several cases,
perhaps in all, had developed into a sac, the wall of which was
made up of ectoderm and entoderm layers.
In the same way on August 2 tissue was allowed to settle over
the bottom of a watch glass, forming a very thin deposit. Great
numbers of small lumps formed which had the same history as the
above. It is quite possible of course that a few of these lumps sent
out coenosarcal outgrowths, but that such escaped notice.
Summary. Eleven experiments with Hudendrium were made.
In all experiments fusion led to the formation of plasmodial
masses. In eight experiments an extensive formation of perisare
took place. In five experiments numerous small masses with
perisare, which remained alive for days, were formed; in three of
BEHAVIOR OF DISSOCIATED CELLS 307
these experiments several of the spheroidal masses sent out
coenosareal outgrowths. In two experiments a considerable
number of coenosareal outgrowths were obtained from nodules
of tissue which had remained alive in comparatively large flat-
tened plasmodial masses, and in one experiment these outgrowths
gave rise to hydranths, four in number.
The first experiment, July 9, was much the most successful.
In those that followed a common error undoubtedly lay in endeay-
oring to handle too much tissue. But over and above that the
water grew warmer with the advancing season and the Euden-
drium colonies perhaps became more abundantly infested with
protozoan ectoparasites.
The technique in general of these experiments is especially
faulty in that, (1) it allows parts which must die, tentacle and gon-
ophore fragments, to get into the cultures; and, (2) it subjects
masses of tissue that evidently need the best environment to the
harmful influences of quiet water in a laboratory dish. Small
gauze floats kept at the top of a running aquarium were used, but
with no success. Very probably if the cultures were placed out-
side in the harbor water, they would do better. As to methods,
the following may be added to what has already been said:
Only clean colonies or parts of colonies should be used. If the
whole hydroid is used I believe that colonies without gonophores
are the best, and those with female gonophores the worst. Stem
tissue would perhaps be better than that of the whole colony.
Care should be taken to allow the cells and small lumps to cohere,
and not to break up the cohering tissue at first more than is neces-
sary. It is well to get the tissue in comparatively small pieces a
few hours after preparation, and in fresh dishes away from the
original surface of attachment. If the masses of tissue a few hours
after preparation appear soft and pasty, they will probably not
live. They should show a good color, absence of the characteris-
tic color indicating presence of dirt or infesting animals. The
gauze (silk bolting cloth), usually used runs 50 meshes to 25 mim.
A cloth running about 75 meshes to 25 mm. was also used. The
sea water was well aérated and filtered. An effort was made to
pick out from the cultures all coarser particles, such as pieces of
30S H. V. WILSON
hydranths and gonophores, that could be seen with the eve. The
tissue was usually pressed out and kept in solid watch glasses with
a cavity 50 mm. in diameter and 10 mm. deep. These were im-
mersed in crystallization dishes 200 to 250 mm. in diameter or in
finger bowls of about 120 mm. diameter. Dishes, instruments,
and gauze were thoroughly cleaned, but were not. sterilized.
Possibly sterilization in hot water would be advantageous. I
lay emphasis on the technique of the experiments in the hope that
it may be improved by others. With a more certain technique
this method of growing hydroids ought to lend itself to the pro-
duction of hydrids, as I have suggested (’O07b, ’11b) for sponges.
Histological study of the restitution masses of Hudendriwm
Observation record, July 27. At 4:30 p.m. a drop of the Euden-
drium tissue was squeezed directly from the gauze sae on to a
slide. A supported cover was put on and slide examined at once.
The fluid contained quantities of separate cells. In addition
there were present a few small masses each consisting of several
cells. All the cells were about spheroidal, but they varied a
great deal in size. Some contained abundant pigment granules,
others a few, and others were quite transparent. Four types
could be distinguished, all of which were abundantly represented,
but plenty of transitional cells connecting these types were also
present. The types are shown in fig. 10, a, 6b, c, d. Cell a is
well filled with pigment granules which appear brownish red by
transmitted light. Cell 6 contains similar granules, but they are
few in number. Cells or particles ¢ and d are transparent and either
without pigment or show only a faint granule or two. The slide
was kept in a moist chamber and examined at intervals for four
hours. The formation of numerous minute masses each consist-
ing of a few cells was observed. These grew large through fusion
with one another from hour to hour. At 8 p.m. a few plasmodial
masses of considerable size were present, the largest of which
measured about 300u x 50u. This was a thin flattened, sheet-like
mass having an irregular outline. Its general body was opaque,
but the peripheral part was thinner and here it could be seen that
all of the four types of cells entered into the composition of the
BEHAVIOR OF DISSOCIATED CELLS 309
mass. At this time, 8 p.m., numerous small plasmodial masses
grading down to aggregates of a few cells were also present. In
many of these too the four types of cells could be distinguished.
Finally in the preparation at this time there still remained quan-
tities of free cells. In this preparation there were no bits of ten-
tacles, gonophores, or foreign particles.
Ci
igh X aN
an
Fig. 10 Eudendrium. Elements of the pressed out tissue. > 700.
Fig. 11 Eudendrium. From a section through a lobe of the mass shown in
fig. 7. cen, enidoblast; f. c., free cell; p, perisare. X 1200.
Results from study of sections. The plasmodial mass shown in
fig. 7 was about twenty-four hours old when preserved. A part of
it had already died, but there were three large lobes of living tissue
surrounded by an obvious perisare. These lobes were found to
agree in structure. Part of a section through one of them is
shown in fig. 11. In the interior of the lobe there are numerous
cells which seem to be free, that is the body is well defined all
round. These cells vary in size; the nuclei are relatively large with
310 H. V. WILSON
abundant nucleoplasm and usually with a conspicuous nucleolus;
the cytoplasm as a rule shows vacuoles and solid inclusions in
vacuolar spaces. Similar cells are met with which are not sharply
delimited all round, but only on one side (cell a in fig. 11); on the
other side, the cell shading off into the syneytial reticulum. In
such a case, I take it, we have a mass which has broken away from
the general syneytium on one side, the protoplasm on this side
condensing to form a film of exoplasm. Cnidoblast cells, ¢ 7.,
with included nematocysts are also common in the interior of the
lobe. Between the cells or groups of cells the protoplasm exists
as a vague reticulum, the vacuolar spaces in which are of all sizes.
Scattered in the reticulum are nuclei. The external stratum of the
lobe in some places is not markedly different from the interior.
In other places it shows smaller cells and more nematocysts than
the interior. No doubt some of the nematocysts have been
carried over from the parent hydroid; possibly others are new for-
mations. The external structure of the lobe is in most places
continuous with the interior, but here and there it is separated
as in the next mass to be described. The dead part of the mass
(fig. 7) consists of a loose granular stuff including some nuclei and
nematocysts. The live and dead tissue are not sharply separated,
but grade into each other.
Two other plasmodial masses about twenty-four hours old were
sectioned. These masses were irregular bodies of the same
general character as the one shown in fig. 7. They were however
entirely alive, and surrounded everywhere by a distinct perisare.
They proved to be essentially alike in internal structure. Part
of a section through one is shown in fig. 12... The body is solid and
an outer stratum is almost everywhere separated from an inner
mass by a vaguely delimited cleft-like space, s. The outer stra-
tum is chiefly composed of comparatively smooth cells, forming
four or five layers, and of enidoblasts. The cells have large nuclei
and are in general well defined. There are places however where
one can only find nuclei lying in a vaguely reticular protoplasmic
matrix The inner mass is a complex syneytium containing
abundant large nuclei and vacuoles with inclusions. Numerous
enidoblasts are seattered through it, and well defined ordinary
BEHAVIOR OF DISSOCIATED CELLS Silt
A
2.
Sa ye pee PS
Die eee
Vig. 12. Eudendrium. From a section through a restitution mass about
twenty-four hours old. cn, enidoblast; f. c., free cell; p, perisare; p. /., protoplas-
mie film; s, space separating an outer stratum, o. s., from the inner mass, f. ™”.
x 1200.
Fig. 13 Eudendrium. Section of the restitution mass, with one coenosareal
outgrowth, shown in fig. 8. Section strikes the original mass and does not include
the outgrowth. y, yolk mass. Other lettering as before. > 350.
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, NO, 3
aH LP H. V. WILSON
cells are found in it here and there. Protoplasmic films, p /,
are common which mark off cells or areas on one side while on the
other side the protoplasmic area has no distinct boundary.
The masses shown in fig. 8, two spheroidal and two with out-
erowths, were sectioned. These bodies when preserved were
three days old. In fig. 13 is shown a section through the dilated
body of one of the masses which had a coenosareal outgrowth.
All the bodies proved to be in essentially the same condition as
far as the differentiation of layers is concerned. In them all an
ectoderm and entoderm are distinctly differentiated. The two
layers are separated by a cleft-like space, there being no distinct
supporting lamella. In the small spheroidal masses and in the
dilated portions (representing the original shape) of those with
outgrowths, there is a central yolk that is still continuous with the
entoderm in spots. The yolk mass does not extend into the out-
erowths. The perisare is laminated and in the sections almost
everywhere widely separated from the ectoderm. The ectoderm
is composed of small cells, probably all interconnected, forming
in places one layer, in other places two or three layers. Small
nematocysts and stages in development of these are common,
and other inclusions also are present in the ectoderm. The
entoderm consists of a single layer of large closely packed cells
varying from a more or less cubical to a somewhat flattened shape.
These cells are uninucleate, the eytoplasm more or less vacuolated
and sometimes containing small (developing) nematocysts and
other inclusions. The central yolk, y, is a granular mass in places
composed of small spheres of varying size. In it small spheroidal
vesicles containing one or two deeply staining granules are
common, doubtless representing swollen and degenerating nuclei.
A few small nematocysts are also found in the yolk. The yolk
mass although in general separate from the entoderm is perfectly
continuous with this layer in spots.
While my observations on the histological strueture of the
restitution masses, both in Eudendrium and Pennaria, are frag-
mentary, they are nevertheless definite. From them it would
seem that in Eudendrium the solid aggregate formed by the fusion
of the isolated cells passes into the condition of a syneytium which
BEHAVIOR OF DISSOCIATED CELLS 313
includes partially or perhaps completely free cells (fig. 11). An
outer stratum in which cell bodies are well differentiated, and
which is several layers deep, now becomes marked off from an
inner mass (fig. 12). The outer layer probably represents the
ectoderm, while the inner mass represents a yolk-entoderm, which
subsequently splits into the definitive entoderm and the yolk.
Finally ectoderm, entoderm and a central yolk mass appear (fig.
13), as in the development of a coelenterate planula, Manicina
for instance (Wilson, ’SS8), or Eudendrium itself (Hargitt, 04).
PENNARIA. RESTITUTION FROM DISSOCIATED CELLS
Species used. The species used was Pennaria tiarella MceCrady
(Proc. Elhott Soc., vol. 1, no. 1, p. 153; Nutting, ’01, p. 337).
Pale specimens with light colored ova and deeply colored speci-
mens with orange ova were abundant together on the floats
round the laboratory wharf at Beaufort during August. Both
pale and colored forms liberated medusae at dusk, about 7 p.m.
The forms appear to represent merely the extremes in ‘a range of
color variation (for the varieties at Woods Hole mde Hargitt, 00).
Experiment July 26. vigorous and clean colony was pressed
out in the usual way at 4:25 p.m. Quantities of cells of various
kinds, especially spheroidal granular cells with more or less pig-
ment, came through; also small cell groups, and bits of tentacles.
Fusion of the cells and cell aggregates begins at once and proceeds
rapidly. In ten minutes time a mass 500ux 100u has been
formed, in a watch glass kept under the microscope, practically
from isolated cells. Such masses change their shape slowly and
fuse with one another. The tissue which was pressed out in a
watch glass was shaken to the center at 4:35 p.m. At 4:53 it has
formed a thin coherent cake. This is now sucked up in piaces
with the pipette, and so broken into pieces about 51min. in diameter
which are transferred to a bowl of water.
At 6 p.m. the tissue lies on the bottom of the bowl in the
shape of thin, somewhat reticular sheets. These are freed from
the bottom (they had already begun to curl up round the edge)
with small pipette and are transferred with a large pipette to a
314 H. V. WILSON
fresh bowl. In so doing the sheets are of course broken where-
upon the peripheral parts turn white and disintegrate quickly.
But the body of the piece remains alive, keeps its color (a reddish
tinge), contracts and soon has a comparatively smooth surface
once more. Fraements accidentally broken from the sheets are
Fig. 14 Pennaria. Restitution mass, four days old. Photograph. ™& 50
also transferred, and these also quickly ‘heal.’ All the masses
continue to contract, and by 8 p.m. many of them have a massive
shape, although some at this time are still sheet-like. They all
have a smooth surface, and the majority of them are in the neigh-
borhood of 1 mm. in diameter. One of the largest is shown in
fig. 14.
BEHAVIOR OF DISSOCIATED CELLS yl lPa)
On July 27 at 9 a.m. the masses are surrounded by a distinct
perisare and with some exceptions are still alive or include con-
siderable live tissue. The color of the live tissue is pink. The
smaller masses are of compact shape, spheroidal or ovoidal, and
are alive throughout. The larger masses are of a somewhat
lobular shape, and while the projecting lobules are alive, a consid-
erable part of the body of the mass is dead or dying. Several are
now preserved. Sections confirmed what f have just said as to the
distribution of the live tissue. In fig. 20 a section through one of
the smaller masses is figured, and it may be seen that the whole
mass was alive, and while in general still solid had begun to ditfer-
entiate the ectoderm and entoderm layers. The masses at this
stage are soft and burst easily on rough handling. The water
was henceforth renewed with a siphon.
On July 29 two of the masses have developed outgrowths. A
photograph of one of them is shown in fig. 15, and the other is
represented by a photograph (fig. 16) and a camera sketch (fig.
21), the latter made from the living object. The original mass in
both cases was spheroidal, and the thick perisarc, 0. p., which
surrounded it, still persists. The mass, as sections of similar
bodies show, has differentiated into ectoderm and entoderm layers
which surround a central cavity containing the remains of yolk
material. The body shown in fig. 21 has developed one long
outgrowth ¢ and two short ones, a and 6, just protruding at the
opposite end from the original perisare. In the long outgrowth
the ectoderm and entoderm are thicker than elsewhere. The peri-
sare over this outgrowth is noticeably thinner than that over the
original mass, and the ectoderm in a part of the outgrowth has
contracted away from the perisare, remaining connected with it
by strands, ect. s. after the fashion characteristic of the adult
hydroid. The original mass, too, it may be seen has contracted
away from the perisare and has materially changed its once globu-
lar shape. The other mass, fig. 15, which has developed only a
single outgrowth, represents a slightly earlier stage than figs.
16 and 21. In it the original mass has contracted away from the
perisare, 0. p., but remains connected with it by strands of ecto-
derm. In the outgrowth however the ectoderm has not yet
316 H. V. WILSON
1.5, 16
Fig. 15 Pennaria. Restitution mass three days old. a, coenosarcal out-
growth; op, perisare of original mass Photograph Xx 50
Fig. 16 Pennaria Restitution mass three days old. op, perisare of original
mass; p, perisare of long coenosarcal outgrowth. Photograph x 50
begun to separate from the perisarcal covering. Both these
masses including the outgrowths are firmly adherent to the bot-
tom of the vessel.
On July 29 a large number of the masses are dead. Only the
smaller ones together with two or three of the larger survive, and
the latter have been injured and evidently are in bad condition.
Injury often comes in changing the water, the siphon setting up a
current which strains the bodies of the larger masses especially
since these are attached to the bottom only throughout a part of
their extent. The small masses are more perfectly attached to
the glass. It is clear that if one wishes to grow hydroids in this
way it is better to produce comparatively small masses instead of
large ones such as that shown in fig. 14.
BEHAVIOR OF DISSOCIATED CELLS
17 18
Fig. 17 Pennaria. Restitution mass, five days old, with two outgrowths,
one branched op, perisare of original mass Photograph x 5O
Fig. 18 Pennaria Restitution mass, two days old, metamorphosed, with
hydranth. op, perisarc of original mass. Photograph x 50
On July 30 one of the masses has developed three outgrowths,
rach about like the long outgrowth in fig. 16. The mass itself
and two of the outgrowths adhere to the bottom, while the third
rises obliquely in the water. On July 31 only one of these out-
erowths remains adherent to the bottom; the other two rise
obliquely in the water. The extremities of the latter have now the
character of knobs, reddish brown in color and resting upon
lighter colored stalks. This mass continues to develop and on
August 1 has reached the condition shown in fig. 19. The ascend-
ing outgrowths now bear hydranths, each with the lower filamen-
tous tentacles and the upper short capitate tentacles characteristic
of this hydroid. The thick perisare, 0. p., marks out the size of
the original mass from which the outgrowths sprouted. The out-
erowth x is the one that remained adherent to the glass. In the
318 H. V. WILSON
Fig. 19 Pennaria. Restitution mass six days old, completely metamorphosed,
with developed hydranths. op, perisare of original mass; x, perisarc of outgrowth
adherent to glass. Photograph. % 50.
living body the perisare ended in a closed rounded extremity
and contained a coenosareal prolongation extending throughout
its length. When the body was pried from the glass, the coeno-
sareal prolongation retracted and in the photograph it appears
very short.
Experiment August 3. A clean vigorous colony about 5 inches
high is selected and only stem material (coenosare) is used. All
lateral branches are cut off, also the base and tip of the main
stems. The latter are then cut into pieces about 3 mm. long and
BEHAVIOR OF DISSOCIATED CELLS 319
Fig. 20 Pennaria. Section of restitution mass, seventeen hours old. ec
ectoderm; p, perisarc; y. en., yolk entoderm. X 150.
Fig. 21 Pennaria. Same mass as in fig. 16. a, 6, c, coenosareal outgrowths;
are of original mass; p, perisare of long coeno-
ec, ectoderm; en, entoderm; op, peris
sarcal outgrowth. 90.
Fig. 22 Pennaria. Median section of restitution mass two days old. a, b,c,
short coenosarcal outgrowths; ec, ectoderm; en, entoderm; op, perisare of original
mass; y, yolk material. > 150.
320 H. V. WILSON
these are pressed through gauze in the usual way. The tissue
thus obtained is pure coenosareal tissue broken up into separate
cells and minute aggregations of cells. The tissue is pressed out
at 8:10 p.m. Fusion goes on rapidly, and in twenty minutes
round the edge of the collection of tissue small bars and plates, |
to 3. mm. long, have been formed. At 4:50 p.m. the material
exists as a thin cake about 4mm. thick, of considerable coherency,
vet of loose reticular texture. This les on the bottom scarcely
adherent to the glass. It is cut with scalpel and needle into pieces
about 4 mm. in diameter. Nine such pieces are prepared and
together with some much smaller fragments are transferred with a
pipette to a fresh bowl of water.
At 7 p.m. all the pieces have contracted considerably. The
plate-like pieces have begun to curl up at the edge and the smallest
fragments have already assumed a compact, massive shape. The
color is a light pink. On the next morning (August 4) all the
masses are still alive and have secreted a distinct perisarc. The
smaller are spheroidal or ovoidal in shape, the larger of an irregu-
larly massive shape, measuring up toa length of 3mm. They are
all adherent to the glass. The water is now changed by siphoning,
and some of the larger masses rupture. They rupture very easily
and it is clear they are too large to thrive. After rupturing, a
mass quickly acquires once more a smooth surface within the
perisare.
On August 5, two of the masses have partially transformed.
One is shown in section in fig. 22. This mass was originally
spheroidal and in the living state, including the perisare, meas-
ured about 4 mm. in diameter. It was firmly attached to the
elass, and at 9 a.m. August 5 was still spheroidal. At 2 p.m.
it was observed to be triangular. The triangular character be-
‘ame more marked during the afternoon, and it was plain that
the mass was sending out three outgrowths, a, b,c. The body was
preserved at 7 p.m. Seetions showed that the originally solid
plasmodial body had differentiated the ectodermal and entodermal
layers.
The other mass on August 5 had developed farther. A photo-
graph of this mass is shown in fig. 18, and a camera sketch made
BEHAVIOR OF DISSOCIATED CELLS yA
from the living object in fig. 23. The original mass was spheroi-
dal; its outline is indicated by the thick perisare, 0. p. Two short
outgrowths, a and 6, had developed, and these together with the
original mass adhered to the glass. A third outgrowth c¢ ascended
in the water and had transformed into a hydranth bearing whorls
of short stubby tentacles. The ectoderm and entoderm had
developed throughout the mass, and the ectoderm of the tentacles
and of the short outgrowths included abundant nettle cells. The
hydranth while under the miscroscope, in a watch glass, was
frequently active, bending from side to side. The size of the
gastric cavity varied with the contraction state but during most
of the time was large. Shortly after transferring the mass from
the breeding dish to the watch glass and just after fig. 23 was made
a quantity of granular material was twice ejected from the mouth,
after which the polyp contracted considerably. The water was
then changed, and the polyp returned to about the condition
shown in the figure. It was then preserved (4 p.m.). The
tentacles of the hydranth arranged in three whorls, all look alike,
and are of a more or less globular shape. The upper whorls
doubtless represent the capitate tentacles of the adult. Possibly
the tentacles of the lower whorl elongate and develop into the
filamentous tentacles. In the development of the egg polyp,
Hargitt (O00, p. 400) finds that the filamentous tentacles appear
first, the capitate somewhat later In the restitution polyp
shown in fig. 23 there may have been a slight difference in the
time of appearance of the several whorls.
On August 6 another small mass has developed an outgrowth
and resembles fig. 15. On the next day it is dead. By this time,
August 7, many of the masses including all the larger ones are
dead, but some survive and of these three have developed each an
outgrowth. One of them is shown in fig. 24, the others are sub-
stantially like it. In all three the coelenterate layers have re-ap-
peared; the original mass has contracted away from the perisarc
with which it remains connected by ectodermic strands. While
the mass shown in fig. 24 was under the miscroscope, these strands
retracted into the body. The three partially transformed masses
are now preserved.
aZZ H. V. WILSON
23 24
Fig. 23) Pennaria. Samemassasinfig.18. a,b, short coenosarcal outgrowths;
c, outgrowth that has become a hydranth; ec, ectoderm; en, entoderm; m,
mouth; op, perisare of original mass; ¢, tentacle. X 90.
Fig. 24 Pennaria. Restitution mass three days old, with outgrowth. Let-
tering as before. X 90.
On August 8 only four of the masses of this experiment are alive.
Two are still spheroidal. One has developed an outgrowth and
is substantially like fig. 24. The other is shown in fig. 17; it has
given rise to two outgrowths at opposite poles, one of which has
developed a lateral branch. The original mass has contracted
away from the perisarc, remaining connected with it by ectoder-
mic strands in the usual way. On the perisare algae have settled
and these appear in the photograph as rounded spots. In all
of these bodies, even in the spheroidal masses, sections showed that
the ectoderm and entoderm had developed, and a gastral cavity
containing the remnants of a yolk mass was present. The bodies
were preserved August 8.
Summary. Two experiments were made. In the first both
stems and hydranths were used. In the second only stem tissue
was used. A large number of solid plasmodial masses were
obtained, and these within a day uniformly secreted a distinct
perisare. In the first experiment all the larger masses gradually
died. Three masses } to 3 mm. in diameter, differentiated the
BEHAVIOR OF DISSOCIATED CELLS 323
coelenterate layers and developed coenosarcal outgrowths. One
of these masses went further and developed perfect hydranths.
In the second experiment also the larger masses died. Ten smaller
masses for the most part about 3 mm. in diameter, the largest
reaching a diameter of :‘) mm., differentiated the ectoderm and
entoderm layers. Of these, eight developed coenosarcal out-
growths, and of the eight one mass produced an actively motile
hydranth with whorls of tentacles.
Comparison with egg development. For the purpose of compari-
son with the restitution hydroids my assistant, Mr. O. W. Hyman,
reared Pennaria from the egg. As Hargitt (00) states, the eggs
vary considerably in size, the planulas in size and shape both.
Several planulas were measured and it was found that they
ranged in length from about :> mm. to 1 mm., while the cross
diameter was about 7 mm. It will be seen that these planulas
were not so far removed in bulk from the restitution masses that
transformed, although neither in the case of the planulas nor in
that of the restitution masses was there any uniformity of size.
On the other hand the hydranths produced by metamorphosing
planulas are fairly constant in size, and they agree in this respect
with those produced by the restitution mass shown in fig. 19.
Histological study of the restitution masses of Pennaria
Sections of the Pennaria stem show that the entoderm is made
up of a single layer of large columnar cells tapering towards the
base and measuring about 30u by l5u. The cells contain very
many large spheroidal granules that stain pale blue with haema-
toxylin. The ectoderm contains an abundance of nettle cells,
large and small. In each layer the elements are freely inter-
connected to such a degree that in many regions at least. the
structure is that of a reticular syneytium.
In order to study the composition of the pressed out tissue |
squeezed pieces of stem in a watch glass of water so that the
squeezed out tissue fell on an immersed cover glass. After five
minutes strong formalin was added. Much of the tissue thus
fixed adheres to the cover, and this when mounted on a slide in
water gives clear pictures. In such a preparation, fig. 25, there
324 H. V. WILSON
Pig. 25 Pennaria. Elements of coenosarecal tissue. From a preparation
fixed five minutes after tissue was pressed out x GOO.
are quantities of large granular cells, many of which are spheroidal
as a, other with pseudopodia as c, some with a larger vacuole as
b. The granules are generally scattered all through the cell,
but as in d there may be some clear protoplasm. These cells
exist separately but also in small aggregations as d and e. The
cells and contained granules are of about the same size as those
seen in sections of the stem entoderm and the cells evidently
represent the entodermic elements. In the preparation again
are numbers of cnidoblasts of various sizes with included
nettle cells, /. Other elements, doubtless also eetodermic, are
finely granular pale cells varying a good deal in size,.g, to-
gether with small groups of such cells. When the stem piece is
pressed, numbers of cells must be ruptured, and the preparation
contains quantities of free granules, b, doubtless derived in large
part from the entoderm cells. These are transparent, and for the
most part spheroidal, although often irregular in shape. They
seem to stick together and even to fuse. Droplets of translucent
substance smaller than the entodermie granules and varying in
size down to the vanishing point are common. It seems probable
that some of this material spoken of as droplets and granules
represents minute fragments of protoplasm that have rounded off.
And the question is worth formulating, although I can not answer
it: are such minute granular or drop-like bodies, representing por-
BEHAVIOR OF DISSOCIATED CELLS 325
tions of broken down cells, incorporated in the restitution mass as
it grows? Again in such a preparation one finds some masses, 7,
composed of finely granular material, with cell boundaries here
and there vaguely showing, and sometimes with included nettle
cells. These are probably lumps of ectoderm.
If such a preparation be made and examined alive in sea water,
the same elements are observed. Formalin removes the color
from the granules in the entoderm cells, but in the living prepara-
tion it may be seen that they have in general an orange tint,
ranging from yellow to reddish, although some are colorless. The
entoderm cells execute slow amoeboid changes of shape.
When the stem pieces are pressed through gauze and the tissue
is at once examined alive in a drop of sea water, it is found to con-
sist of the same elements described above. With this treatment
more entoderm cells seem to be ruptured and there are fewer
groups of cells. The coenosare is almost entirely broken up into
the elements a, /, g, h, of fig. 25. When the entire hydroid is cut
up and pressed through gauze, again the same elements are found
if the tissue is examined at once, although possibly more aggre-
gates of cells occur than when stem tissue alone is used.
If the drop of live tissue pressed out through gauze, or squeezed
out without using gauze, be kept under observation, it may be
seen that small masses are soon formed which include entoderm
cells, enidoblasts, and the pale cells that probably are of ecto-
dermic origin. As these masses grow in size 1t becomes impossible,
owing to their opacity, to study their composition while alive
As stated already, fusion between the cells of the pressed
out stem tissue goes on so rapidly that in twenty minutes times
small bars and plates, 1 to 3 mm. long, can be drawn off with a
pipette. Some of these were preserved and sectioned and it
could be seen that such masses were solid bodies of fairly uniform
structure showing no stratification into incipient layers. The
superficial part does not differ from the interior. The structure
throughout is that of a cellular syncytium, that is in certain regions
no cell boundaries can be seen, the protoplasm here appearing
as a syncytial mass containing scattered nuclei, while in other
places cell boundaries are visible. Even where cells are marked
326 H. V. WILSON
out it is probable that they are interconnected with one another
and the rest of the mass by protoplasmic strands. The cells
that are marked out vary in size and shape. Here and there cells,
usually in groups, may be recognized by the contained granules
as the original entoderm cells. Only a small fraction of the mass
however is now made up of such cells, and yet the entodermic
elements composed a very large part of the tissue when the fusion
masses began to form. It is plain then that the entoderm cells
after fusion to form, or rather help form, the plasmodial masses,
undergo a transformation which effectually precludes us from
recognizing them later. The large cnidoblasts formed a conspicu-
ous set of elements in the tissue when fusion began, and these are
to be seen in very considerable numbers scattered throughout
the plasmodial mass at the stage under examination (twenty min-
utes old). A comparison between the sections of this and later
stages indicates that the bulk of the nematocysts carried over
from the parent gradually disappear during the development of
the plasmodial mass. If this is so, it is a question of some inter-
est what becomes of the enidoblast cell itself? Does it share in the
formation of the regenerative tissue? The point is worthy of
special study, including as it does the idea of the de-specialization
of a highly differentiated element.
The protoplasm of the cell and syncytial areas in this stage is for
the most part finely vacuolated so as to present a reticular appear-
ance to a high power. The nuclei are in general large and con-
tain abundant nucleoplasm. The mass at this time seems to
have no surfacé film apart and distinet from the superficial syney-
tial and cell areas. Finally in connection with this stage it may be
said that owing to the transformation which the entoderm cells
undergo after fusion, it does not seem hopeful to attack from
purely histological evidence the question as to whether ectodermic
and entodermic elements become segregated, the ones on the out-
side, the others in the interior of the mass. There is of course
always a possibility that this occurs, but it seems remote.
Somewhat older plasmodial masses formed by the fusion of
stem tissue pressed through gauze were studied. These were
preserved 1 hour after fusion had begun, and were considerably
BEHAVIOR OF DISSOCIATED CELLS 327
larger than the mass just described. Sections show however that
they have essentially the same structure. Cells or protoplasmic
areas with the entoderm granules are still recognizable here and
there. Perhaps most of such cells lie in the interior but some are
found at the surface. This is not a point of importance for the
question as to the possible segregation of ectodermic and entoder-
mic elements, since as I have already explained a very large part of
the entodermic material can no longer be recognized as such. It
is clear that many of the areas or elements of the plasmodial mass
now without granules, and with a vacuolated protoplasm, must
have been formed from entoderm cells.
The solid plasmodial mass does not long remain unstratified.
As already said the masses uniformly secrete a perisare within
about one day after fusion begins, and this in itself is probably
evidence that the superficial layer has assumed something of an
epithelial character. Masses preserved July 27, about seven-
teen hours old, were sectioned. Some of them were healthy and
alive all through, and fig. 20 represents a section through such an
one. In this body an outer layer, ec., the ectoderm, has separated
from an inner mass, y. en., the yolk entoderm. The latter as
later stages show gives rise to the definitive layer of entoderm and
a central yolk mass, as in the planula development of Pennaria
(Hargitt, 00, 04). Both ectoderm and yolk entoderm are cellular
syneytia in which free elements or apparently free elements are
included. In both the ectoderm and yolk entoderm, some of the
large nematocysts carried over from the parent, are present. On
one side of the body it will be seen the differentiation into layers
has not yet been carried out, the ectodermal region here shading
off into the inner mass. The isolated and irregular cavitiesin the
yolk entoderm doubtless represent the beginnings of the gastric
cavity. Inothermassesof the same lot preserved at the same time
the shape was lobular, and only the projecting lobules were alive,
while the more central part of the body was dead or dying. See-
tions showed that in the lobules the layers were present, in about
the same stage of differentiation as in fig. 20.
Restitution masses from Experiment August 3 that were pre-
served nineteen hours after fusion began were sectioned. These
proved to be in about the condition shown in fig. 20.
THE JOURNAL OF EXPERIMENTAL ZOOLOGY. VOL. Il, NO. 3
328 H. V. WILSON
A mass two days old, from Experiment August 3, was sectioned,
and a median section is represented in fig. 22. The mass was orig-
inally spheroidal, but gave rise to three short outgrowths a, ), c.
In this body the ectoderm and entoderm are well differentiated.
The yolk entoderm of the earlier stage has obviously given rise
to the entoderm, en., and to more centrally located yolk material,
y, and the latter has been nearly absorbed. In places the ento-
derm is still continuous with yolk elements. The best sections
show that the entoderm is still a reticular syncytium, but the cell
bodies are distinctly outlined in places. In the most distinct
regions they have a columnar shape as in the adult. The ento-
derm is now well stored with the spheroidal granules found in the
adult cells. The ectoderm also appears to be in reality a reticular
syncytium, but cell bodies are clearly differentiated and regularly
arranged in the regions of the outgrowths. They have here an
elongated, columnar shape. Some large nematocysts, apparently
such as were carried over from the parent, are present in the ecto-
derm. A very few such nematocysts are found in the entoderm,
and these seem to be in a phase of dissolution.
Other masses from the same experiment (August 3) that were
four and five days old were sectioned, and the results may be
briefly given. While the mass is still spheroidal and before it
has developed outgrowths, it may differentiate an ectoderm, ento-
derm and central yolk. By the time a well defined layer of ento-
derm is present, the yolk mass is small in amount and consists of
scattered spheres or small groups of spheres. In such spheroidal
masses the ectoderm and entoderm have the character of reticular
syneytia. As outgrowths develop, both ectoderm and entoderm
assume the character of columnar epithelia, especially in the out-
growths themselves.
LEPTOGORGIA. FUSION OF DISSOCIATED CELLS
The species used was Leptogorgia virgulata, the ‘sea feather’
that is common, especially under piers, in Beaufort harbor. Some
introductory experiments were made under my direction by my
assistant, Mr. O. W. Hyman. He established the fact that when
Leptogorgia is cut into small pieces, and these pressed in gauze sacs
BEHAVIOR OF DISSOCIATED CELLS 329
in the usual way, the tissue is broken up into small masses and
separate cells. By the subsequent union of such masses and cells,
smooth balls up to and over 1 mm. in diameter are formed. ‘These
remain alive for days in laboratory dishes, but do not transform.
A record of two subsequent experiments is here given.
Experiment August 9. Pieces 4 to 5 mm. long are cut from the
upper end of a yellow colony. The horny axis occupies about
one-fourth the total diameter. These pieces are simply squeezed
with forceps in a watch glass of water. The tissue exudes freely
from the cut ends. Much of it is stringy. With pipette it is
dispersed and so broken up. It is then shaken to the center of
the watch glass and the glass transferred to a bowl of water, 11:15
a.m.
Some of the tissue is now examined under a supported cover.
It is made up as follows: (1) Ciliated cords and masses, vary-
ing in size from large to minute are abundantly present.
These are doubtless pieces of mesenterial filaments with mesen-
terial tissue. (2) Motionless masses of loosely packed cells,
also varying considerably in size are abundant. (3) Isolated
cells and groups of a few cells are abundant (fig. 26). In fig. 26,
a represents a characteristic small cell group made up of a few
spheroidal cells with sharp outlines, some full of highly refractive
granules, some merely containing a good many such. Similar
granules form a compact mass at one end of the cell group, but this
mass lacks a bounding pellicle. Separate spheroidal granular
cells, b, resembling the constituents of a are common, and they
may have pseudopodia. There is an abundance of small spheroi-
dal masses of glassy protoplasm, c, the larger with one or two
granules. Finally there are plenty of isolated granules, d, such
as are found in the granular cells. In the category c the smallest
elements must surely be fragments of cells or bits of intercellular
connectives that have rounded off.
The preparation under the microscope was watched for about
an hour, and it was observed that fusion took place involving all
the classes of constituents above enumerated. At 12 m. several of
the larger ciliated masses were motionless, the surface bearing
instead of cilia numerous small pseudopodia and transitional
stages from cilium to pseudopod.
330 H. V. WILSON
The tissue left in the bowl was examined at intervals. The
masses. grew evidently through fusion with one another and
through incorporation of the granular cells and other elements.
A typical mass at 2 p.m. is shown in fig. 27. The mass is full
of granules like those of fig. 26, and in it the outlines of some sphe-
roidal cells, like fig. 26 b, are distinguishable. Round it are similar
granular cells and very many small masses of glassy protoplasm
some with a few granules, some without any, ranging in size from
mere points up nearly to the diameter of the granular cells.
By 9 p.m. the process of fusion had gone so far that spheroidal
masses with smooth surface, from about } mm. in diameter down-
wards, were present. A number of these were now picked out and
transferred to fresh sea water. In the formation of these masses
about ten hours transpired. It is evident that during this time
some regressive differentiation of the fusing lumps of tissue to a
simpler condition took place. What the character of these
changes were and how uniform they were, it would be interesting
to know.
The further history of the masses is briefly as follows. On
August 10 they were alive. Atthistime they are perfectly opaque,
and the smooth surface shows abundant fine flagellum-like
pseudopods. Many of them are surrounded by a deposit of whit-
ish material. This on examination proves to be made up of
spheroidal cells of various sizes and degrees of granulation, which
evidently have been slowly given off from the mass. This giving
off of cells continues during the next day. It is probably evidence
of a bad condition of the mass, and it is noteworthy that some of
the masses do not exhibit it.
All of these bodies were preserved on August 11, and several
were later sectioned. While still alive something could be learned
of their histological structure by gently crushing them under a
cover glass. On doing so some of the contents streams out in the
shape of small spheroidal masses. The larger of these are like the
granular cells (6) of fig. 26. From such they range down to mi-
nute particles of glass-like protoplasm just large enough to be seen
at a magnification of 600. Of the intermediate sizes some are
full of granules, others contain one or a few, while still others are
BEHAVIOR OF DISSOCIATED CELLS 331
Fig. 26 Leptogorgia. Cell aggregate and free elements of pressed out tissue.
xX 1200.
Fig. 27 Leptogorgia. Fusion mass two hours fifteen minutes old. > 150.
without granules and glass-like. The granules in all these bodies
are mostly of one size and yellowish. Such an examination by
no means necessarily implies that in the natural condition the
mass is composed of spheroidal cells. Rather, I assume that, on
crushing, the cell bodies are separated and, where such exist, the
intercellular strands are broken. A quick contraction would then
make all the protoplasmic masses spheroidal. A body that has
been slightly erushed under a cover glass in this way may heal.
One such was kept for two hours in a moist chamber, and at the
end of this time the body healed perfectly and was once more
surrounded, as it originally was, by a smooth surface pellicle.
Sections showed that these bodies did not have a uniform com-
position. In one (ball 1) the structure was as follows: There is a
surface film but sections give nothing definite as to its composi-
tion. The interior is solid and shows no stratification into layers
(ectoderm and entoderm). In many regions one finds protoplasm
studded with nuclei but no cell boundaries can be made out.
In other places there are small cells, rounded or angular that are
very closely packed. In still other places while the tissue is
compact the cells are slightly separated by unstained substance,
probably fluid. In such places the cell bodies are distinctly
outlined and intercellular strands of protoplasm are freely present.
3a2 H. V. WILSON
In a few small areas there is a scanty accumulation of mesogloeal
jelly imbedded in which are some strands and small masses of
nucleated protoplasm which are freely interconnected. The
jelly stains blue with haemalum (or haematoxylin) and with a 2
mm, objective appears homogeneous. These several observations
show that the mass is a syneytium in different parts of which cell
bodies are differentiated in various degrees.
Fig. 28 Leptogorgia. Median section of a fusion mass two days old. Layers
interpreted as ectoderm, ec, and entoderm, en, have developed. ms, mesogloeal
jelly. > 350.
Another mass (ball 2) has the structure shown in fig. 28. | There
is a surface film which appears as a mere line. The general mass
consists of the syncytial cellular tissue described for ball 1, but
round half of the body two layers, apparently the ectoderm and
entoderm, are differentiated. These layers are distinetly differ-
entiated, although there is no mesogloeal jelly between them. At
about the middle of the body they fade away into the general
mass. The entoderm consists of more or less columnar elements,
BEHAVIOR OF DISSOCIATED CELLS BaD
the ectoderm of irregular cell bodies separated by a good deal of
fluid and freely interconnected. Near the center of this ball is a
considerable collection of mesogloeal jelly, of the same character
as that deseribed for ball 1.
In still another case (ball 3) while a part of the body resembles
ball 1, at one end of the body the structure is that shown in fig.
29. We find here an accumulation of finely granular, lightly
stained material, cg, which is quite different from mesogloeal jelly
and is probably a fluid that has coagulated in the fixation process.
Fig. 29. Leptogorgia. Part of section through a fusion mass two days old,
cg, coagulum; s.c., superficial cells. > 1200.
The same ball contains some mesogloeal jelly near the center. In
the region round the coagulated fluid the cells are loosely packed.
They are more or less rounded and intercellular connections are
practically absent. Vacuoles are common in these cells and a con-
spicuous nucleolus is frequently to be seen. In this region the
surface layer is formed of flattened cells, s.c. In this ball there is
no differentiation of ectoderm and entoderm layers.
Interpreting the results of this study of sections, it seems prob-
able that ball 1 represents an earlier stage, and ball 2 a later one
in which the coelenterate layers have begun to differentiate.
334 H. V. WILSON
In ball 8 the condition shown in fig. 29 is perhaps to be correlated
with the gradual extrusion of rounded granular cells which goes
on in the case of some of the masses during life, as already recorded.
Such a condition is one, perhaps, into which the dense syncytial
cellular structure passes when the struggle for life is going against
the body. It may of course only be a mortuary change, viz., a
step in a process of gradual dying.
The results of this experiment, while very inconclusive, suggest
that the bodies formed by the fusion of cell masses and cells would
regenerate into new individuals if placed under good conditions,
possibly hung out in gauze bags in a part of the harbor where
the current is good.
BHeperiment August 10. Small pieces of a Leptogorgia colony
were pressed out through gauze at 4.30 p.m. The tissue that
streams through the gauze is finer than that obtained in the experi-
ment just recorded. It is made up of the elements shown in
fig. 26 and of small opaque lumps of tissue, mostly spheroidal and
many of them ciliated, which are commonly three or four times the
diameter of one of the granular cells (fig. 26 6). The living tissue
and the spicules are separated as far as possible.
Fusion goes on and by 7 p.m. masses of irregular shape are
present. These are transferred to fresh sea water. The next
morning a number of smooth balls have been formed, some of
which have incorporated spicules. These are kept for a couple of
days during which they show no external signs of differentiation.
Sections showed that these balls had essentially the same structure
as those of the preceding experiment.
Experiment to test the regenerative powers of a fusion mass when
inserled in the body of the parent species. August 10. Six of the
Leptogorgia fusion masses produced in the experiment of August
9 were inserted in the parent species in the following way. A piece
of an orange colored Leptogorgia, about five inches long, was slit
lengthwise down to the horny axis. The slit so made was pushed
open and the fusion masses dropped in with a pipette in a row.
Ties were then made round the piece of Leptogorgia closing up the
covering layer of polyps over or partly over the fusion masses.
On August 11, 10 a.m., the ties are removed. The slit has not
healed but the edges have curled in. The whitish fusion masses
BEHAVIOR OF DISSOCIATED CELLS aT)
may be seen in the slit. They have fused with one another in
some degree, the number of masses now being four. The piece
of Leptogorgia looks healthy; the polyps are well expanded.
At 7 p.m. the whole preparation is preserved. The questions
are: Have the fusion masses undergone any histological differen-
tiation? Have they established union with the Leptogorgia?
Sections through one of the masses showed that the body had
grown deeper into the slit and had established connection with the
Leptogorgia on one side of the slit. This connection included per-
feet continuity with the entodermie lining of a coelenteric cavity
which had been laid open, and also with the entoderm of several
small coenenchymal canals in the neighborhood. The whole
fusion mass is solid and somewhat club-shaped at its outer end
where it shows a stratification into an outer stratum and an inner
core. The thickness of the outer stratum is considerable includ-
ing several layers of cells. The other masses did not penetrate
so deep into the slit. They were found to be in continuity with
the superficial layer of the Leptogorgia, but had not established
connection with the interior of the latter.
The indication from this experiment, which was merely meant as
a tentative one, is that the fusion masses if allowed to grow would
have become part of the Leptogorgia colony.
ASTERIAS. FUSION OF THE DISSOCIATED CELLS OF THE
IMMATURE GONAD
Experiment August 5. Gonads 25 mm. long of the common
starfish, Asterias arenicola, were cut into pieces about 5 mm. long
and these were pressed through gauze at 12m. Abundant sepa-
rate cells and small cell aggregates stream through the gauze to-
gether with some larger pieces of gonad. The latter are picked
out, and the remaining material is shaken to the center of watch
glass. A drop of the material is now examined under the micro-
scope. Many of the cells whether free or combined in small aggre-
gates are coarsely granular. Both cells and aggregates show
fine pseudopodia. The cells and the aggregates quickly combine
and in a few minutes the field of the microscope presents the
appearance shown in fig. 830. There are numerous small masses,
such as a, which have been formed by the fusion of cell aggregates
¥510) H. V. WILSON
and separate cells; and there is an abundance of free elements.
Among the latter coarsely granular cells like 6, with and without
pseudopodia, are conspicuous. There are also many clear glass-
like cells (¢) ranging down to bits just visible. Free granules
resembling those of the granular cells are abundant. The masses
(a) make the impression of being aggregations of the granular
cells (b), but doubtless other elements enter into their composition.
Abundant fine pseudopodia, occasionally branched, cover the sur-
face of such masses.
By | p.m. the tissue in the watch glass has combined to form
a thin and extensive reticular plate, produced by the gradual
a
Fig. 30 Asterias. Small fusion mass and free clements of pressed out gonad
tissue. a X 600. b and e X 1200.
fusion of masses of many shapes. The reticulum in general is
attached, though feebly, to the glass, but pieces 1 to 2 mm.
wide have been broken off and are free. At 7 p.m. the whole
reticulum is broken up into pieces of about this size, and all
transferred to fresh sea water. On the following day a number of
such pieces had contracted into smooth, massive bodies. But
all pieces died in a day or two.
BEHAVIOR OF DISSOCIATED CELLS Bad
Experiment August 16. Gonads 20 mm. long were used. The
gonads were cut into pieces and these simply teased up with needles
in a watch glass of sea water. The pieces are thus broken into
small masses, cells, and fragments, essentially like those shown in
fig. 30. Fusion commences at once as in the former experiment,
the isolated cells and the masses both throwing out pseudopodia.
The granular cells in particular are observed to become intercon-
nected by delicate and complex pseudopodial networks. The
formation of pseudopodia by the granular cells may go on to such
an extent that the granular substance of the cell body almost
disappears in the network of pseudopodial strands. This e peri-
ment was not carried farther.
Sections showed that the gonads used in these experiments were
in the indifferent stage. The germinal epithelium lining the
follicle is more than one layer deep, and many of the nuclei are
large and rich in chromatin. The epithelium has proliferated to
such an extent that the lumen is nearly filled with cells. In
sections these are compressed, with rather vaguely granular
eytoplasm and nuclei which are smaller than in the lining cells.
When the living gonad is slightly pressed, these cells exude and
appear as the spheroidal granular elements described above.
LITERATURE CITED
Braem, I. 1908 Die Knospung der Margeliden, ein Bindegled zwischen
geschlechtlicher und ungeschlechtlicher Fortflanzung. Biol. Central-
blatt, Bd. 28.
Cuinp, C. M. 1906 The development of germ cells from differentiated somatic
cells in Moniezia. Anat. Anzieger, Bd. 29.
1911 Die Physiologische Isolation von Teilen des Organismus als
auslésungsfaktor der Bildung neuer Lebewesen und der Restitution.
Vortrige u. Aufsiitze a. Entw-Mech. d. Organismen, Heft 11.
Curtis, W. The life history, the normal fission and the reproductive organs of
Planaria maculata. Proc. Boston Soe. Nat. Hist., vol. 30.
Drinscu, H. 1902 Studien wt. das Regulationsvermégen der Organismen:
6. Die Restitutionen von Clavellina lepadiformis. Arch. f. Entw.-Mech.
Bd. 24.
Evans, R. 1899 The structure and metamorphosis of the larva of Spongilla
lacustris. Quart. Journ. Mier. Sci., vol. 42.
338
Harairr,
LIEBERKU
Maas, O.
Metscunt
H. V. WILSON
C.W. 1900 A contribution to the natural history and development
of Pennaria tiarella MeCr. Amer. Naturalist, vol. 34.
1904 The early development of Pennaria tiarella MeCr. Arch. f. Entw.-
Mech., Bd. 18.
HN, N. 1856 Beitrage zur Entwickelungsgeschichte der Spongillen.
Archiv fiir Anat. u. Physiologie, J. Muller.
1901 Die Knospungentwicklung der Tethya und ihr Vergleich mit
der geschlechtlicher Fortpflanzung der Schwimme. Zeitschr. f. wiss.
Zool., Bd. 70.
1906 Ueber die Einwirkung Karbonatfr. Salzlésungen auf erwachsene
Kalkschwiimme u. auf Entwicklungsstadien derselben. Arch. f.
Entw.-Mech. d. Organismen, Bd. 22.
1910 Ueber Involutionserscheinungen bei Schwiimmen u. ihre Bedeu-
tung fiir die Auffassung des Spongienkérpers. Festschr. z. sechzig-
sten Geburtstage Richard Hertwigs, Bd. 3.
KOFF, E. 1879 Spongiologische Studien. Zeitschr. f. wiss. Zool., Bd. 32.
Mituier, Kart 191la Versuche it. die Regenerationsfihigkeit der Siisswasser-
NUTTING,
PERKINS,
SCHULTZ,
WELTNER,
schwimme. Zool. Anzeiger, Bd. 37, Nr. 3-4.
1911b Beobachtungen tiber Reductionsvorgiinge. bei Spongilliden,
nebst Bemerkungen zu deren diusserer Morphologie u. Biologie. Ibid,
Nr. 5.
C. C. 1901 The hydroids of the Woods Hole Region. Bulletin of
the U. 8. Fish Commission for 1899.
H. F. 1902 Degeneration phenomena in the larvae of Gonionema.
Biol. Bulletin, vol. 3.
E. 1904 Ueber Reduktionen: 1. Ueber Hungererscheinungen bei
Planaria lactea. Arch. f. Entw.-Mech., Bd. 18.
1906 Ueber Reduktionen: 2. Ueber Hungererscheinungen bei Hydra
fusea L. ITbid., Bd. 21.
1907 Ueber Reduktionen: 3. Die Reduktion u. Regeneration des
abgeschnittenen Kiemenkorbes von Clavellina lepadiformis — [bid.,
Bd. 24.
1908 Ueber umkehrbare Entwickelungsprozesse u. ihre Bedeutung fiir
eine Theorie der Vererbung. Vortrige u. Aufsitze iiber Entwicklungs-
mechanik d. Organismen, Heft 4.
W. 1893 Spongillidenstudien II. Archiv f. Naturgeschichte,
Jahrg. 1893, Bd. 1.
Witson, H. V. 1888 On the development of Manicina areolata. Journ. of Mor-
phology, vol. 2.
1894 Observations on the gemmule and egg development of marine
sponges. Ibid., vol. 9.
1907a A new method by which sponges may be artificially reared.
Science, vol. 25, June 7, 1907.
1907b On some phenomena of coalescence and regeneration in sponges.
Journ. Exper. Zool., vol. 5.
1911la On the regenerative power of the dissociated cells in hydroids.
Proc. Amer. Soc. Zoologists, Science, Mar. 10.
1911b Development of sponges from dissociated tissue cells. Bulletin
of the Bureau of Fisheries, vol. 30.
RHYTHMS IN. THE REPRODUCTIVE ACTIVITY OF
INFUSORIA
LORANDE LOSS WOODRUFF anno GEORGE ALFRED BAITSELL
Sheffield Biological Laboratory, Yale University
THIRTEEN FIGURES
In a study of the life history of Paramaecium caudatum by
pedigree cultures, Calkins clearly illustrated the cycle, while the
rhythms in the division rate were later emphasized by Woodruff
in a study of the life history of several species of hypotrichous
Infusoria. The fluctuations in their rate of reproduction were
classified as follows:!
“A rhythm is a minor periodic rise and fall of the fission rate,
due to some unknown factor in cell metabolism, from which recov-
ery is autonomous.”
A eycle is a periodic rise and fall of the fission rate, extending
over a varying number of rhythms, and ending in the extinction
of the race unless it is ‘rejuvenated’ by conjugation or changed
environment” (ef. fig. 1).
Gregory, in a study of the life history of Tillina magna, stated
that ‘The curve which represents the general vitality of the proto-
plasm shows the normal rhythmic fluctuations observed by Wood-
ruff.” Gregory also made an analysis of the data secured by
Popoff in his study of the life history of Stylonychia mytilus,
and she stated that “If the curve of Stylonychia is plotted from
average records of five and ten day periods, it will be found to
correspond to the curves of Paramaecium, Oxytricha and Tillina,
each showing the rhythmic periods of high and low vitality.”
More recent work has shown that Paramaecium aurelia may be
bred indefinitely on a culture medium which is varied from day
to day,*1.e., the cycle does not occur under these conditions though
1 Woodruff (’05). 2 Cf. Woodruff (‘11a), Taf. 26, 27.
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, No. 4
NOVEMBER, 1911
339
F (GO, ‘YNApOOAA) = ‘sauT] UayoIq [Borjtoa Aq poywusSisop oe sumyyAYI
SNOLMBA OY} JO S}IUIT] oY, “Spoyjour ainq[no saistped JepuN vidosnjul JO a}vd UOISsYy ayy UT suUIyZAYA, puw SopOAD, OYVAYSN][T OF, “wOTy
-BNUITYS [BIOyTyAe Aq apoAo ysay Jo pus yw ,poywuoanfey, “WNIpour oanjjno uoIsnjur Avy ‘“sporsod fipp wa} 10} posvsoav uresv ‘syeurrue
JO SOUT] INO} OY} JO UOTSIAIP Jo 99¥A ATIVp oFvIOAV VY} Burmoys “y oiny[no ‘xvypey VyoUyAXG Jo Aaoysty OFT oy} Jo Ydvay T “By
TO6T
“AON pue
f “09q ‘po
‘idy pus £061 Avy pus
Arne = oung pus Avy “IBN ‘qoq puv-use ‘90d JAON ‘pO “4 qsnany Apne oun “dy “Ivy P
098 008 002 009 008 (
00
oO
OF
RHYTHMS IN THE INFUSORIA 341
100 ; 200 288
June 1904 July Aug. Sept. Oct. Nov.
Fig. 2. Graph of the life history of Gastrostyla steinii showing the average
daily rate of division of the four lines of animals which compose the culture, again
averaged for ten day periods. Hay infusion culture medium. To illustrate a
case where the rhythms are apparently absent for a considerable period. Compare
with fig. 3. (Woodruff, ’05)
the rhythms persist undiminished; and also that the same result
may be attained by a constant culture medium of beef extract’
(cf. fig. 4).
It is obvious from these investigations that the life history of
Infusoria in pedigree cultures comprises many minor rhythmic
fluctuations in the fission rate from which recovery is autonomous.
The results with beef extract as a constant medium for Para-
maecium aurelia naturally led to an intensive study of therhythms,
in order to determine if these also can be eliminated by a still
more constant environment, i.e., whether they are due to minor
variations in the environment or to unknown intracellular phe-
nomena, as originally stated. Possible sources of variation in the
environment which might give rise to variations in the metabol-
ism of the cell which would become apparent as rhythms in
the rate of reproduction are: 1) Chemical composition of the
culture medium, 2) Quantity and quality of the bacterial flora
of the culture medium, 3) Excretion products of the paramaecia,
4) Mechanical stimulation during isolation, 5) Light, 6) Baro-
metric pressure, and 7) Temperature. The data secured which
bear on this question are given in the present paper.
8’ Woodruff and Baitsell (711).
GEORGE A. BAITSELL
WOODRUFF AND
LORANDE L.
342
(co, ‘YNapoo Ay) ‘potsed
Jo}IOYS B 1Oj oFv1VAV UB Aq MOIA 0} 4YZNoIG oq uvo puv yuasoid ore Aoyy SsajoyjsoAou ‘spotsad fipp wa} IOJ PASBIIAB ST
o}BI OY} WOYAM ouITy JO YASus] urezt0d vw doy avodde you AvuT suTYyYAYA YSnoyyye yeyy yowy ayy aVVISN][ OF, “spotsed finp
aay 1oy atay poyjoyd vyeq °Z “BY UL UMOYS TTUT4s VTAJSOLStY) Jo AN4[Nd oUIES OY} Jo L1048TY ast] oy} Joydery ‘Sy
JaquIoAON 1940309 Joquieydag qsnany Ange FO6T ‘une
* 886 002 oor
= aa
|
|
|
00
|
— “0
|
OT
—- + — —+-—-+4
J oe
RHYTHMS IN THE INFUSORIA 343
METHODS
The animals employed in this study were taken from the pedi-
gree culture of Paramaecium aurelia (1) which one of us‘ has had
under daily observation for fifty-one months and which has
attained 2500 generations, up to the present time (August 1,
1911), under the conditions of a varied environment, without
conjugation or artificial stimulation. From this culture a sub-
culture was isolated line by line on October 1, 1910, at the 2012th
generation, and carried for ten months on a constant culture
medium of beef extract. It was then discontinued.’ The average
daily rate of division of the four lines of this subculture (IB),
again averaged for five day periods, was computed and the result
is graphically shown in fig. 4.
The experiments in regard to the rhythms were begun on June
8, 1911, by isolating two subcultures line by line from IB at the
2335th generation, and placing the animals in a similar manner
on depression slides in five drops of the beef extract medium.
This medium consisted of a 0.025 per cent solution of Liebig’s
extract of beef. The slides were kept in small moist chambers to
prevent evaporation. The cultures were continued by isolating
each day an organism from each of the four lines of the respective
cultures, and placing it in fresh medium on a sterile depression
slide. The number of divisions during the previous twenty-four
hours was recorded at the time of isolation and from this data the
graphs were drawn. One of these two subcultures was placed in
a thermostat chamber at a temperature of practically 82° F.
(culture IB82a) and the other in a chamber at a temperature of
practically 76° F. (culture [B76a), and maintained at this tempera-
ture for forty days.
A second series of two subcultures was similarly started from
IB on June 18th, at the 2346th generation, and treated exactly
the same as the above cultures. The cultures of this series were
designated IB82b and IB76b, respectively. A third series of
two subcultures was isolated in the same manner from IB on
June 28th, at the 2355th generation, and these cultures were
. 4 Woodruff (’11a).
5 Wor further details of this subculture (IB) ef. Woodruff and Baitsell (’11).
344
gg ok
LORANDE L.
WOODRUFF AND GEORGE
r=}
—
1 CY —
BYuNLvesadW3l SNOISIAIG
A.
BAITSELL
Part 1
4
345
«
a
IN THE INFUSORIA
RHYTHMS
‘ystsiod sumyyAyI oy} ssofayjioaou ‘padoydura st 4oviyx9 Joaq Jo WMIpeul 91n4
-]NO JURYISUOD aTqRIIMS B WAY IN990 4OU seop efoAd 9Y} YSNOYIS yey JoRJ oY} oyBAYSNI[I OF, ‘spotsod Sutpuodsesi09
ay} 1OF (*,) eangvsoduie} WOOL oSvIOAV ayy SoAIS ydvis waddn oy, ‘spotsed Avp aay 10} padvisav urese ‘ainyj[nod sty}
ASIdUIOD YOTYA S]J9O ,1O4SIS, JO SOUT] INO} oY} JO WOTSTATIp Jo ay¥A ATIVp oBvIOAG OY} YUaSeIdar SoyvUIpPIO oY, “WNIpeu
ainjpNd yowryxo Joog “TT6T ‘6g A[NE UO ponurTzUOdsIp sv 41 [GUN ‘OT6T ‘T 19q099Q UO JT a1N}]Nd WIOIZ UOTYBOTSI S4I JO aUATy
ay) Woy ‘Gy orngqnoqns ‘eIpaine wNTOsBUIBIe JO A1oysty ajeTduI0D oy} Seats ydvIs amo] oY, (Zpue syed) FSI
GWed—V
346 LORANDE L. WOODRUFF AND GEORGE A. BAITSELL
designated IB82c and IB76c. There were then the following
cultures, all of which were kept in the dark, involved in this
experiment:
IB—at room temperature.
IB82a—isolated from IB and continued for forty days at 82° F.
1B76a—isolated from IB and continued for forty days at
IB82b—isolated from IB and continued for forty days at 82° F.
IB76b—isolated from IB and continued for forty days at 76° F.
IB82c—isolated from IB and continued for thirty days at 82
IB76c—isolated from IB and continued for thirty days at 76°
Since the entire point involved in this study depends upon the
constancy of the environment to which the animals are subjected,
this will be considered in detail.
1. Chemical composition of the culture medium. The culture
medium was made up by weighing out the proper amount of
Liebig’s extract of beef and diluting it with distilled water. The
solution was then put into about one hundred test tubes, plugged
with cotton and sterilized. The medium remained sterile until
used. Since all the culture medium which was used throughout
the experiment was made up at one time there was no variation
in the medium itself during the work.
2. Quantity and quality of the bacterial flora of the culture medium.
Paramaecium i$ an animal which depends on bacteria for its food,
and consequently these must be supplied. Sufficient bacteria
were ‘automatically’ transferred with the animals at the first
isolation to provide ample food until the next isolation at the end
of twenty-four hours. Again at this time sufficient bacteria
were ‘automatically’ carried over with the animals to infect the
fresh medium in which they live for the following twenty-four
hours, and so on. The quality of the bacterial flora was initially
the same on all the slides because all the paramaecia used to start
the various lines were taken from the same environment when the
experiments were begun, and it is believed that this condition was
maintained by the cross infection of all the slides almost daily.
This also served to eliminate variations due to infections from
the air of the moist chambers. Obviously the number of bacteria
on a slide varied during the twenty-four hours between isolations.
RHYTHMS IN THE INFUSORIA 347
But a study of the preparations showed that the paramaecia
keep down the results of the rapid multiplication of the bacteria
by feeding on them, so that, although there is ample food for the
animals at all times, the variation in the bacterial content of the
medium is not so great as would at first glance appear to be the
case. However, the point to be emphasized is that these varia-
tions, small as they were, were only of twenty-four hour duration
since fresh culture medium was supplied daily. Consequently
any effect of the slight and unavoidable variation in the quantity
of the bacteria could result only in an intradiurnal rhythm in
the division rate, and since the count of the generations was taken
at twenty-four hour intervals, this variation would not appear in
our records. Elaborateness of method is not necessarily coin-
cident with exactness of technique, and therefore it was considered
unnecessary to attempt to ‘sterilize’ the paramaecia and feed
them on pure cultures of bacteria. Any effort in this direction
has met only with partial success and has introduced compli-
eating factors which, it is believed, would more than counter-
balance any advantages to be gained for the problem in hand.
3. Light. Throughout the experiments all the cultures were in
absolute darkness except for the short time daily when the count
was being made. This was unavoidable, but each animal was
not exposed to the light for more than three minutes. A control
culture carried in the light showed that light does not influence
the rate of reproduction of paramaecium. This is in accord
with the previous results® on the effect of light on the division
rate of free-living Infusoria.
4,and 5. Excretion products of paramaecia and mechanical
stimulation during isolation. These factors may be eliminated
because they could only give rise to an intradiurnal rhythm
which would not appear in the data.
6. Barometric pressure. A careful study was made of the varia-
tions in the barometric pressure which occurred during the experi-
ments. There was absolutely no correlation between the small
fluctuations in pressure and the rhythms in the division rate, and
§ Maupas (’88) and Woodruff (’05).
348 LORANDE L. WOODRUFF AND GEORGE A. BAITSELL
consequently it can be positively stated that this factor plays no
part in our results.
7. Temperature. This is the chief possible variable in the envir-
onment which we have to consider. In the original discussion of
rhythms it was stated that ‘‘the results serve to emphasize the
fact that while temperature does influence the rate of multipli-
cation, it is not the most important element among the factors
which cause fluctuations.” A study of figs. 4 and 5 shows that
there is a certain amount of correlation between the fluctuations
of fission rate and of temperature, when the cultures aresubjected
to the ordinary changes in temperature of the laboratory.
Our experiments were carried on in a Panum thermostat,’
heated at one end with a gas flame (with an automatic regulator)
and cooled at the other by a large ice chest. The thermostat
was divided into nine chambers, grading down in temperature
from one end to the other of the apparatus. The temperature
was recorded in each chamber by a maximum and minimum
registering thermometer, by a tube thermometer, and in one
chamber also by a thermograph. Experiments were conducted
in six of the nine compartments, but an account is given here of
the results of the cultures at the two temperatures within the
optimum zone for the strain of Paramaecium being used. The
detailed data in regard to the effect of different temperatures on the
division rate of this animal and its relation to the temperature
coefficient of chemical reactions in general will be published later.
We should state, however, that our results at other temperatures
are entirely concordant with those here described.
The temperature selected for the study of the influence of
temperature on the rhythms were approximately 82° F. and 76° F.
as it was found that the optimum zone for the culture included
these points. During the fifty days that the experiments covered
the variations in temperature did not exceed 3° F.; for the greater
part of the time the variation was less than 1° F., and for several
periods it was less than 0.5° F. The greatest variation noted
7 This apparatus was constructed for this and similar studies from an improved
model designed by Professor L. F. Rettger of the Sheffield Bacteriological Labora-
tory of Yale University. Our thanks are due Dr. Rettger for his kind codperation.
|
i}
|
OXYTRICHA ~ B
|
PLEUROTRICHA = A
PLEUROTRICHA =~ B
}
Dec. Jan. February March April
1902 1903
Fig. 5 Sections of the culture graphs of two cultures of Oxytricha fallax and
two cultures of Pleurotricha lanceolata, together with the graph of the average
room temperature for the same period. Averages for ten day periods. To illus-
trate a striking instance of the apparent relation of rhythms to the fluctuations in
temperature. (Woodruff, ’05)
349
350 LORANDE L. WOODRUFF AND GEORGE A. BAITSELL
above occurred during a week of unusually hot weather when the
sudden change was too great to be immediately compensated for
by the automatic regulator. Great care was taken in removing
the preparations from the chambers for the daily count and isola-
tion. The culture medium to be used on one day had’ been put
the day before in the proper chamber of the thermostat, and con-
sequently the animals were transferred to fresh culture medium
of the same temperature. Of course any effect of variations
arising from the daily transfers could only be intradiurnal and con-
sequently would not appear in our results. It should also be
emphasized that the recorded temperature was that of the air
in the thermostat, whereas the animals were in culture fluid on
slides within the moist chambers in the thermostat. The tem-
perature of the moist chambers obviously was still more constant
than that of the thermostat chamber, as likewise was the liquid
in which the organisms were living. Consequently the varia-
tions in temperature which the animals experienced certainly
never exceeded 3° F. throughout the experiments, and this maxi-
mum variation occurred only at one period. Fora period of ten
consecutive days there was no visible variation greater than 0.4° F.
It is believed that the temperature conditions were maintained
as nearly constant as modern apparatus and the necessities of the
experiment allowed.
RESULTS
The results can be stated briefly because graphs of the rate of
reproduction bring out the points involved far better than a
description by words. Fig. 6, A, gives the average daily rate of
division of the four lines of ‘sister’ cells of Paramaecium aurelia,
series IB82a, again averaged for ten day periods, at 82°F. Band
C show the same for series IB82b and IB82c._ Fig. 7, A, B, and C,
shows similarly the results derived from IB76a, and IB76b, and
IB76e. Figs. 8 and 9 give the same data averaged for five day
periods. Fig. 10 shows the average daily rate of division, for
five day periods, of line 1 (of the four lines) of series IB82a and
line 1 of IB76a. Fig. 11 gives the same data for series IB82b
and IB76b. Fig. 12 presents the average daily rate of division
RHYTHMS IN THE INFUSORIA Sol
A B Cc
Fig. 6 A, Graph of the average daily rate of division at 82°F. of the four lines
of ‘sister’ cells of Paramaecium aurelia, series IBS82a, again average for ten day
periods. B and C, Similar graphs for series IB82b and IB82c respectively. To
illustrate rhythms in the fission rate when the cultures are subjected to practically
constant conditions, including temperature.
A B &
Fig. 7 A, Graph of the average daily rate of division at 76°F. of the four lines
of ‘sister’ cells of Paramaecium aurelia, series IB76a, again averaged for ten day
periods. B and C, Similar graphs for series IB76b and IB76c respectively. To
illustrate rhythms in the fission rate when the cultures are subjected to pract ically
constant culture conditions, including temperature. Compare with fig. 6.
302 LORANDE L. WOODRUFF AND GEORGE A. BAITSELL
A B Cc
Fig.8 A, Graph of the average daily rate of division at 82°F. of the four lines of
‘sister’ cells of Paramaecium aurelia, series IB82a, again averaged for five day
periods. B and C., Similar graphs for series IB82b and IB82c respectively. To
illustrate the fact that rhythms in the rate of division appear more pronounced
under practically constant environmental conditions. Compare with the last
ten periods of the culture subjected to room temperature changes (fig. 4).
of the four lines of IB82a and IB76a, respectively. The vertical
dotted lines include the ten day period during which temperature
variations were entirely absent, or not greater than 0.4° F. Fig.
13 gives the same results for series IB82b and IB76b during the
ten days of most constant temperature.
A study of these graphs of the rate of reproduction of Para-
maecium shows that the exceptionally and practically constant
conditions of the environment failed to diminish or eliminate the
rhythms—but on the contrary tended to bring them out more
clearly. The fact that the rhythms appear more pronounced
A B Cc
Fig.9 A, Graph of the average daily rate of division at 76°F. of the four lines of
‘sister’ cells of Paramaecium aurelia, series IB76a, again averaged for five day
periods. B and C., Similar graphs for series IB76b and IB76e respectively. To
illustrate the same point as fig. 8.
RHYTHMS IN THE INFUSORIA 353
Fig. 10 Graph of the average daily rate of division for five day periods of line
1 (of the four lines of ‘sister’ cells) of series IBS2a (= continuous line) and of
line 1 (of the four lines of ‘sister’ cells) of series IB76a (= - - - - line). To
illustrate the fact that rhythms of practically the same amplitude and character
appear in a graph of a single line of cells as appear when four such lines are
averaged together. Compare with fig. 8, section A, and fig. 9, section A.
under the practically constant conditions existing during these
experiments than they do under ordinary laboratory conditions,
clearly suggests that they are due to a fundamental factor in
cell phenomena and not to extraneous causes. For if they are
due to inherent intracellular conditions, one would a priorz expect
to find them more clearly brought out when the cell is free from
extraneous influences.
A study of the curves of the division rate at the two tempera-
turesshows that temperature, as is well known, markedly influences
the rate, but it also shows that the rhythms persist—the repro-
ductive activity being, as it were, pitched at a higher scale, but
its character in no wise altered. In other words, it is not sug-
354 LORANDE L. WOODRUFF AND GEORGE A. BAITSELL
(ees)
Fig. 11 Graph of the average daily rate of division for five day periods of line
1 (of the four lines of ‘sister’ cells) of series IB82b (= continuous line) and of
line 1 (of the four lines of ‘sister’ cells) of series IB76b (= - - - - line). To
illustrate the same points as fig. 10. Compare with fig. 8, section B, and fig. 9,
section B.
gested that the division rate is not largely a function of tempera-
ture—all other conditions being equal. It is probable that the
temperature coefficient of the mean rate of division for a period
including several rhythms will coincide closely with that of chem-
ical reactions in general, but it is also probable for example that
the rate of division at the crest of a rhythm at a high temperature
and at the bottom of a rhythm at a low temperature will give a
coefficient higher than the theory demands. Experiments to
determine this point are in progress.
It should also be pointed out that the total number of divisions
attained during a prolonged period of time is comparatively con-
stant. For example, the number of generations attained by
culture I during 1909 was 613 and during 1910 was 612. Of
ud
RHYTHMS IN THE INFUSORIA
*SUOT}IPUOD oINyVIoduIE, yURYSUOD AT[BOTJOvAd Jo potted v BulIMp o}eA oy} UT SUOT}
-BIIVA POYIVU OY} MOYS 0} PUR ‘UOISTATP JO 9}¥1 ATIVp OY} JO plodad [BoTdAY B aYBAASNT[L OT, “YoG°O UBY} Sse] 1M Popslodor
SUOTIVIVA 9INYVIOdUId} oY} YIM Surmp sAvp wo} oy} opNjour soul] peqyop [BoTyA0a oy, *(EUTy
(ouT] snonutyu0s =) BZF
J] Set1as JO S|]
90
‘
IO4SIS, JO SOUT] ANOJ oY} JO UOISTATP JO o}vA App dBvIOA
---- =) BOJqI jo pus
ayy jo ydeay
a
G
1 Sl
4
NO.
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL, 11,
356 LORANDE L. WOODRUFF AND GEORGE A. BAITSELL
Big. 13 Graph of the average daily rate of division of the four lines of ‘sister
cells of series IBS2b (= continuous line) and of IB76b (= - - - - line) for the ten
days during which temperature variations were practically absent. To illustrate
the same points as fig. 12. Compare with the section of fig. 12 included within the
vertical dotted lines.
course this very exact coincidence is an ‘accident’ but, taken with
a considerable amount of data along the same line, it quite defi-
nitely points to the fact that the organism has the potential for
about a certain number of bipartitions during a long period of
time and this number is approximately attained irrespective of
the minor fluctuations in the rate, due to external or internal
causes.
In a recent paper, Jennings states that
“Within the same line the rate is sometimes very different for a cer-
tain period, as a week or ten days, from the rate during the rest of the
time. This is much more evident when one inspects a table in which the
fissions are recorded day by day. The rate in a given line is there seen
at times to drop, remain low for perhaps ten days, then return to the
RHYTHMS IN THE INFUSORIA 357
original rate. In most of all these cases there are evidences of patho-
logical conditions during these periods of lowered rate of fission. Mon-
strosities appear, and many of the specimens die. Therefore these peri-
ods of slower rate are not to be considered as giving the characteristic
rate for the race when healthy. In comparing different races, the
periods when the rate of fission is high and uniform should be compared.”
These observations of Jennings are difficult to understand in
view of our results with Paramaecium. His statement in regard
to weekly variations we would, at first glance, interpret as further
evidence of rhythms; but throughout the more than four years
of the life of this pedigree culture, a monster has never been seen
in any of the direct lines, and only two or three times has a single
deformed individual been seen in the heavy stock cultures which
have been seeded from this strain. Further, it is an unusual
occurrence for a line in any of our experiments to die out without
assignable cause. Therefore it is necessary to emphasize that
whereas the statement quoted seems, at first thought, to be in
regard to periodic fluctuations in the rate of bipartition identical
with those we call rhythms, nevertheless these fluctuations have
absolutely nothing in common since, according to Jennings’
statement, those occurring in his lines are pathological.
CONCLUSIONS
The results of studies on the life history of free living Infusoria
by exact pedigree culture methods show that, when these organ-
isms are bred on comparatively constant culture media of hay
or other infusions, the reproductive activity shows cycles and
rhythms. Further results show that when Paramaecium aurelia
is bred on a varied culture medium, or on a constant medium of
beef extract, cycles do not occur, but rhythms persist. The
results given in the present paper show that it isnot possible by
constant environmental conditions to eliminate the rhythms and
to resolve the graph of the multiplication rate into an approxi-
mately straight line. It therefore seems justifiable to conclude
that there are inherent rhythmical changes in the phenomena of
the cell which are brought to view still more clearly when not
358 LORANDE L. WOODRUFF AND GEORGE A. BAITSELL
influenced by external factors. Variations in the rhythm of divi-
sion is well-known in the development of the metazoon egg and it
has yet to be satisfactorily explained. Towle in a paper on the
effects of stimuli on Paramaecium makes the following state-
ment: “There may even prove to be rythmical changes in sensi-
tiveness like those described by Lyon for cleaving eggs, and Scott
for unfertilized eggs. Something of this nature is indicated by the
fact that paramaecia from the same culture vary in sensitiveness
from day to day.”’ Woodruff (05) wrote: ““In my work on the
effect of chemicals on Infusoria I have found that individuals
react differently at various times to a given stimulus and I believe
we have the clue to these changes in sensitiveness manifested in
the rhythms of the fission rate.”’
Finally, the data justify the conclusion that the cells of this
pedigree culture of Paramaecium aurelia have the potentiality
to perpetuate themselves indefinitely by division (under proper
environmental conditions)—the only necessary variations in the
rate of reproduction being normal minor periodic rises and
falls of the fission rate, due to some unknown factor in cell
phenomena, from which recovery is autonomous.
LITERATURE CITED
G.N.Carxins 1904 Studies on the life history of Protozoa. IV. Death of the
Aseries. Conclusions. Journ. Exper. Zool., vol. 1, no. 3.
L. H. Gregory 1909 Observations on the life history of Tillina magna. Journ.
Exper. Zool., vol. 6, no. 3.
H.S. Jennines anp G. T. Harairr 1910 Characteristics of the diverse races of
Paramaecium. Journ. Morph., vol. 21, no. 4.
BE. P. Lyon 1902 Effects of potassium cyanide and of lack of oxygen upon the
fertilized eggs and embryos of the sea urchin. Amer. Journ. Physiol.,
vol. 8, no. 1.
1904 Rhythms of susceptibility and of carbon dioxide production in
cleavage. Amer. Journ. Physiol., vol. 11, no. 1.
E. Maupas 1888 Recherches expérimentales sur la multiplication des infusoires
ciliés. Arch. d. zool. exper. et gén., 2me ser., T. 7.
RHYTHMS IN THE INFUSORIA 359
M. Pororr 1907 Depression der Protozoenzelle und der Geschlechtszellen der
Metazoen. Archiv f. Protistenkunde, R. Hertwig Festband.
J. W. Scorr, 1903 Periods of susceptibility in the differentiation of the egg of
Amphitrite. Biol. Bull., vol. 5, no. 1.
E. W. Towtse 1904 A study of the effects of certain stimuli, single and com-
bined, upon Paramaecium. Amer. Journ. Physiol., vol. 12, no. 2.
L. L. Wooprurr 1905 An experimental study on the life history of hypotrichous
Infusoria. Journ. Exper. Zool., vol. 2, no. 4.
1909 Further studies on the life history of Paramaecium. Biol. Bull.,
vol. 17, no. 4.
191la Two thousand generations of Paramaecium. Archiv f. Protis-
tenkunde, Bd. 21, 3.
1911b The effect of excretion products of Paramaecium on its rate of
reproduction. Journ. Exper. Zool., vol. 10, no. 4.
L. L. Wooprurr anv G. A. Bairsett 1911 The reproduction of Paramaecium
aurelia in a ‘constant’ culture medium of beef extract. Journ. Exper.
Zool., vol. 11, no. 1.
CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSHUM OF COMPARATIVE ZOOLOGY
AT HARVARD COLLEGE, E. L. MARK, DIRECTOR, NO. 226
THE REACTIONS OF EARTHWORMS TO DRY AND
TO MOIST SURFACES
G. H. PARKER anp H. M. PARSHLEY
The certainty with which an earthworm that is creeping over
a partly moistened surface will avoid dry areas is well known to
students of animal activities. It is the object of this paper to
discuss briefly the character of this response, the location of the
receptors concerned in it, and the nature of the stimulus. The
work was carried out on the common dungworm, Allolobophora
foetida (Sav.), but there is reason to believe that the results
obtained apply equally well to most species of earth-worms.
If a normal worm is allowed to creep over a horizontal sheet
of filter paper that is wet with tapwater excepting for a few spots
and if it is directed by some such stimulus as light toward one of
these dry spots, on reaching the spot, it will usually continue to
creep over the dry surface for a distance varying from a few seg-
ments to half its length, stop, swing its head from side to side,
then draw the anterior part of its body back to the moist region,
and finally proceed to crawl in a new direction over the moist
part of the paper. Of seventy worms put to this test only four
failed to show the series of reactions just described. These four
crept completely across an extensive dry area without showing
the characteristic reaction, but a few days later three of these
worms responded in a normal way, showing that their previous
atypical condition was probably due to some unusual and tem-
porary state. It is, therefore, fair to conelude that Allolobophora
as a rule avoids dry areas.
To ascertain the part of the worm that is stimulated by a dry
surface, several kinds of experiments were tried. To test the
sensibility of the posterior end of the worm, individuals were
made to creep backward by touching their anterior ends slightly
361
362 G. H. PARKER AND H. M, PARSHLEY
and, when thus creeping, were directed toward a dry surface.
Thirteen worms tested in this way made considerable backward
excursions over such surfaces showing, as was to have been
expected, that the posterior end of the worm is not especially
sensitive to dryness. Next, forty-five worms, all of which had
been found to respond normally to a dry surface, were deprived
of their prostomiums and in some instances of an adjacent seg-
ment or two, and were shortly afterwards tested on filter paper.
All crept freely over a dry surface without showing the lateral
movements and the retraction of the head characteristic of nor-
mal worms. The regeneration of the prostomium takes place
in from one to two weeks according to the extent of the injury.
This regeneration was found to restore to the worm its original
sensitiveness to dry surfaces and enabled it to react again ina
typical fashion. There is, therefore, every reason to believe
that the region of the prostomium is the portion of the worm
that is stimulated by dryness.
The terminal surface left after the removal of the prostomium
offers more or less of an obstacle to the ordinary locomotor move-
ments of the worm and to avoid this feature in the experiments,
supplementary tests were made in which the prostomium, instead
of being removed, was anaesthetized. After some preliminary
trials, three anaesthetics satisfactory for this purpose were found;
they were a weak solution of chlorotone, a 35 per cent solution of
magnesium sulphate, and a 1 per cent solution of ether, all aque-
ous. If the anterior tip of a worm is bathed with one of these
solutions for from one to five minutes, the prostomium is found
to remain insensitive to a dry surface for one or more days.
Such anaesthetized worms will creep persistently over a surface
of dry filter paper on which, before anaesthetization, they could
not be induced to advance more than a very short distance. Full
recovery from the effects of the anaesthetic occurs in a day or
two. It is, therefore, clear from this evidence also that the pros-
tomium and possibly some of the adjacent parts of the worm are
the receptive surfaces for this response.
It might be supposed that the greater harshness of dry filter
paper as contrasted with moist filter paper, instead of the simple
REACTIONS TO DRY AND MOIST SURFACES 363
absence of water, was the significant factor in these reactions,
but such is not the case, for, if worms are allowed to creep on
surfaces that remain equally rough whether they are wet or dry,
the same reactions are observed as in the tests in which filter paper
was used. Such surfaces as those of bricks, tiles, ete., present
_these conditions. In trials with the moist and dry surfaces of
bricks results similar to those got on filter paper were obtained.
Moreover worms drew back from a dry smooth brick to creep on
a wet rough one, and from a dry rough brick to creep on a wet
smooth one, showing that the presence or absence of moisture,
and not roughness or smoothness were the significant elements
in these reactions.
From these observations, it is quite evident that the prostomial
region of an earthworm can be stimulated by dryness to such an
extent as to call forth vigorous locomotor responses of a charac-
teristic kind. A moist surface seems to be unstimulating and to
afford merely a condition favorable for the locomotion of the
animal. In this respect the earthworm is the reciprocal of the
human being, for our skin is more receptive to the condition of
wetness than to that of dryness. With us, however, the sensa-
tion of wetness is produced in all probability by a complex of
pressure and temperature stimuli, whereas in the earthworm the
response to dryness is dependent very likely upon a simpler stim-
ulus. This is apparently the selective extraction of water from
the peripheral protoplasm of the worm, a process which is favored
by the eapillarity of the dry surface over which the worm begins
to ereep and is probably dependent chiefly upon evaporation
from the surface of the worm itself. Under such circumstances
the materials in the peripheral protoplasm of the prostomium
must become concentrated and probably initiate stimulation by
undergoing some such change as partial coagulation. Processes
of this kind are not well exemplified in the outer skin of man, but
are more nearly comparable with what occur in our mouths when
by excessive evaporation the oral surfaces become somewhat dry.
AN ATTEMPT TO ANALYZE THE CONSTITUTION
OF THE CHROMOSOMES ON THE BASIS OF SEX-
LIMITED INHERITANCE IN DROSOPHILA
T. H. MORGAN
From the Zoological Laboratory, Columbia University
FOUR FIGURES—COLOR PLATE
In several preliminary notes I have given a brief account of
the origin of four mutations in the eye-color of the fly, Droso-
phila ampelophila. The heredity of these eye-colors may now
be given in full, and the bearing of the results, on sex-limited
inheritance in general, discussed. In addition to the eye-color
data I shall also. describe a few cases in which two other sex-
limited characters have been studied in connection with eye-
color, namely: short proportionate wings and yellow body-color.
A full account of the heredity of these latter two characters will
be reserved, however, for later publication. Here they are
used only in so far as they give an opportunity to study the mode
of inheritance of three sex-limited characters present in one
individual.!
The experiments on Drosophila have led me to two principal
conclusions:
First, that sex-limited inheritance is explicable on the assumption
that one of the material factors of a sex-limited character is carried
by the same chromosomes that carry the material factor fer femaleness.
Seconp, that the ‘association’ of certain characters in inher-
itance is due to the proximity in the chromosomes of the chemical
substances (factors) that are essential for the production of those
characters.
1 The facts here recorded were first announced in a public lecture given in the
Marine Biological Laboratory at Woods Hole, Mass., July 7, 1911.
365
366 T. H. MORGAN
PART I
THE HEREDITY OF RED, VERMILION (OR BRIGHT-RED), PINK,
AND ORANGE EYES
The eyes of the wild fly are dull red, and may be designated
by the letter R. The bright-red eye is vermilion in color and is
indicated by V in the tables.
The pink eye is more translucent than the red eye, but of about
the same general tone. It lacks the dark fleck seen in the red
and vermilion eye when the eye is examined with a lens. This
black fleck changes its position as the lens travels over the eye.
The pink eye, P, is with a little experience easily distinguished
from the other colors, especially in newly hatched flies. When
the fly gets old the eye turns to a brown color very characteristic
of this type of eye.
The orange is the faintest eye color in the series. If the fly
is very small it may be only tinged with orange. If the fly is
large (coming from a well-fed maggot) the orange eye, O, is deep
orange in shade; and without some experience it may be confused
with the pink eye, especially if a mixed culture containing flies
of different ages and sizes is examined. A little experience will
soon make one familiar with the difference between these two
colors.
I do not hesitate to state that there are no intergrades between
these eye-colors. Each color is distinct and breeds true to its
kind. Moreover, the heterozygous flies show the dominant
color. One ‘dose’ is indistinguishable from two doses of the color
determiner. '
In making the matings and recording the numbers I have been
assisted in the experiments with eye-color by Miss Eleth Cattell;
and in the experiments in sex-limited inheritance for three
factors by Miss E. M. Wallace and by Miss M. B. Abbott. I
wish to express here my appreciation of the assistance that they
have given. I have discussed the theoretical results of the eye-
color inheritance with Mr. A. H. Sturtevant, as the work went on.
CHROMOSOMES AND SEX-LIMITED INHERITANCE 367
and this discussion has been helpful to me in finding suitable
formulae for the data.
Rather than defer the discussion of the interpretation of the
results to the end of the account I will take up each case in turn.
A few words will suffice to make clear the symbolism used. The
red eye of the wild fly seems to contain three pigments: red, pink,
and orange. The mutants have arisen by the loss in turn of one
of the factors that make possible the development of the red
color. If these three colors (or the factors that stand for them)
are represented by the symbols R, P, and O, then the red eye is
RPO, the pink eye is rPO, the orange eye is rpO. Obviously
there is another combination possible, viz: the loss of the pink
factor and the retention of R and O, giving RpO, which is the for-
mula for the bright red or vermilion eye. The matter may be
better expressed in another way. Should from any cause what-
soever the factor for pink (P) drop out, vermilion (RpO) would
appear. If, on the other hand, the red factor (R) should be lost
from the red-eyed fly, pink would result (rPO). By crossing a
vermilion fly with a pink one, some orange-eyed flies (rpO)
would appear in the second (inbred) generation by recombination.
In the formulae that follow it is always assumed that one dose
of red or pink gives the same result as do two doses, which accords
with the facts. .
It is necessary to say a word in advance about sex determina-
tion in these flies. I assume that every egg after eliminating its
polar bodies, contains the sex chromosome, called X. Prior to
their extrusion the egg, like all the other cells of the female,
contains two X’s or XX. The male cells contain one X. Half
the spermatozoa contain one X, the other half lack X. Miss
N. M. Stevens has shown that these relations are actually present
in Drosophila. The peculiar ‘coupling’ of X with the factor for
pink, that runs through the formulae and gives the significant
results connected with sex-limited inheritance, will be discussed
later.
368 T. H. MORGAN
Red eye by vermilion eye
When red-eyed females were crossed with vermilion-eyed
males all the offspring (93 in number) were red. These inbred
produced in the second generation red females, red males, and
vermilion males. The result shows that vermilion is sex-limited.
The following table gives the results and numerical data:
Red tQ.ancs stn ont een rE O02
Red @ by Vermilion #7 = fs 2» Fests DE. ee Oe Le)
aoe | Vermilion MAMAS AS aes - 110
The two classes of males taken together number 289, which is a
close approximation to the 302 red females. The results may
be accounted for in the following way:
Red 9°
Vermilion @
= RPOX — RPOX
= RpOX — RpO
F Red @ RPOXRpOX
: Red @ RPOXRpO
: Red @ RPOX — RpOX
Gametes of F, Red ¢ RPOX —RpO
RPOXRPOX Red 92
RNGocernivon RPOXRpOX Red 9
Bh eS RPOXRpO Red 3
RpOXRpO Vermilion @
It will be noted that the red females belong to two classes, one
pure, the other heterozygous: the red male is also heterozygous,
while the vermilion male is pure.
The reciprocal cross, namely, red male by vermilion female
gives red females and vermilion males. In other words the
daughters are like the father and the sons like the mother. This
gives what I eall criss-cross inheritance. When these F,’s are
inbred they give red males and females and vermilion males and
females as shown in the next table.
CHROMOSOMES AND SHEX-LIMITED INHERITANCE 369
’
| Red 9 Bagot sree eee LOG
coy RON NG Reduct: Aone enter er oe
eaters uiacrnil J C
Red & by Vermilion 9 acre ee Veunilion. 9 agi. 207
Vermilion = 186
The two preceding crosses are typical for all cases of sex-
limited inheritance in Drosophila, and for some, perhaps for all,
other cases. They may be summed up in the statement that
where in one combination a character in the grandfather is trans-
mitted to his grandsons alone, the reciprocal combination gives
criss-cross inheritance.
The number of males and of females in each class is approxi-
mately equal in the F, generation. The results are accounted for
as follows:
Vermilion @ RpOX — RpOX
Red & RPOX — RpO
Red 2 RpOXRPOX
Vermilion 7 RpOXRpO
Red 9 RpOX — RPOX
SERGI Osh Vermilion *@ RpOX — RpO
RpOXRpOX Vermilion 9°
F, Generation RPOXRpOX Red 9?
RpOXRpO Vermilion &
RPOXRpO Red &
Red eye by pink eye
The results of this cross have been already published (Science,
1911), but the hypothetical explanation not given. For the sake
of completeness the facts must be restated here. Red female
by pink male gave red male and red female offspring. These
inbred gave in the F; generation 3063 red males and females and
169 pink males and females.
370 T. H. MORGAN
Red 9? |
(Red 2 \ Red of“
\Red @ /Pink 9? |
Pink @f
. 8063
Red 9 by Pinkw= ¢
....169
In this case there is no sex-limited inheritance. An analysis
of the result, based on the same formulae, gives the following:
Red @ RPOX — RPOX
Pink o& rPOX — rpO
Red 9 RPOXrPOX
Red &@ RPOXrpO
Red 9 RPOX — rPOX
Gametes of F; Red o& RPOX — rPOX — RpO — rpO
RPOXrPOX
RPOXRPOX )
3 red 9
rPOXRPOX
rPOXrPOX 1 pink 9
RPOXRpO
F, RPOXrpO 3 red &
rPOXRpO
rPOXrpO 1 pink @
The expectation is three times as many red females as pink
females, and three times as many red males as pink males.
The actual ratio is about 20 to 1, taking the two sexes together.
Thus while all the classes,are represented, and the reds in excess
of the pinks, they are much more numerous than expectation.
The cause of this deficit in the pinks will be discussed later when
other similar results can be brought forward.
The reciprocal cross, red males and pink females, gave in the
first generation red males and females. These produced in the
F, generation red males and females, and pink males and females
in the proportions shown in the next table.
CHROMOSOMES AND SEX-LIMITED INHERITANCE oil
Red @ |
_ J Rede Red af
~ \Redi /Pink 2 |
Pink wf
Red & by Pink
The analysis of the result, based on the same formulae, is as
follows:
Pink 9 rPOX — rPOX
Red & RPOX — RpO
Red @ rPOXRPOX
Ha Red @ rPOXRpO
Red @ rPOX — RPOX
COURS 1) Red @ rPOX — RPOX — ipO0 — RpO
rPOXrPOX Pink @
rPOXRPOX Red @
RPOXrPOX Red 9
RPOXRPOX Red 9
z rPOXrpO Pink &
rPOXRpO ted
RPOXrpO Red &
RPOXRpO Red &
In this combination also the expectation is three red females to
one pink female and three red males to one pink male, while the
realization is about 5 to 1 for all reds versus all pinks.
Red eye by orange eye
When red-eyed females are crossed with orange-eyed males
all of the offspring have red eyes. These inbred produce red-
eyed males and females, pink-eyed males and females, and ver-
milion-eyed, males and orange-eyed males. No females with
vermilion or with orange eyes appear in the second generation.
Here two characters are, in a sense, sex-limited, although the
parents showed only one of them, viz., the orange.
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, No. 4
«
Bie T. H. MORGAN
Red OA Ae. cae eee 138
Rede Gatien. oc 102
, _ J Red @ } Vermilion @.... ; eee OS
Heda ThRvAO Teneo =e aedie PYRO Rs oi -by ee 5
Pinner ccc ote e3
Orange ¢..... Fara eovs Pence: 5083
The number of pinks and oranges is very small, although the
other classes contain a fair number of offspring. The analysis
follows:
Red 2 RPOX — RPOX
Orange o@ rpOX — rpO
F, Red 9 RPOXrpOX
Red & RPOXrpC
Red? RPOX — rPOX — RpOX — rpOX
ees ont Redo’ RPOX — rPOX — RpO — rpO
RPOXRPOX | ;
3
RPOXrPOX red Q
rPOXRpOX |
rPOX1POX 1 pink @
RpOXRPOX |
RpOXrPOX }; 3 red 9
rPOXRPOX
F, rPOXrPOX 1 pink @
RPOXRpO |
RPOXrpO a8} red J
rPOXRpO j
rPOXrpO 1 pink @
RpOXRpO |
RpOXrpO 3 vermilion
rpOXRpO |
rpOXrpO 1 orange o
CHROMOSOMES AND SEX-LIMITED INHERITANCE 373
In this case the expectation was far ahead of the realization;
for, while the red females to the pink females are estimated as
3 to 1 they are as 37 to 1 in the actual count. Again the vermilion-
eyed males should be as numerous as the red males, but they are
not half as numerous. Thus while the formulae give the classes
actually realized in the experiment with the sexes properly dis-
tributed—a matter of no small complexity—yet the numerical
results are by no means the expected ones.
The reciprocal cross, red males by orange females, gives red
females and vermilion males. These inbred produce eight classes
in the second, or F2, generation. These eight classes represent,
in fact, the whole gamut of eye-colors.
Red @ Be jen LOS
REGU Ue eras icra cise ctaseeye LOO
Vermilion @ Epix estechsias 151
Redo pyran pene ui Red oN Vermilion SAG erp OR CH acsntS o 135
(Vermilion #7 SE Pink 9 Soonteieee bee
Pinks Guta eee ti areeeee
[oes OrancepOwa- ee: ae rsasves 13
Oran gevigha nae ees mae cco 31
In this combination the number of pinks and oranges is by no
means so small as in the preceding case, although the other colors
are not much more numerous than before. The analysis is as
follows:
Fi
Orange
Red
Red
Vermilion
T. H. MORGAN
9 rpOX ,— rpOX
o RPOX — RpO
9 rpOXRPOX
ao rpOXRpO
Gametes of Fy
Red @ rpOX — RPOX — RpOX
Vermilion o& rpOX — RpOX — RpO
— rPOX
— rp.O
The expectation in the F, generation is 3 reds, 3 vermilion, 1
The numbers realized are somewhat in this
same ratio, except that the pinks and the oranges again run
pink, 1 orange.
behind their schedules.
rpOXrpOX Orange
rpOXRpOX Vermilion
RPOXrpOX Red
RPOXRpOX Red
RpOXrpOX Vermilion
RpOXRpOX Vermiblon
rPOXrpOX Pink
rPOXRpOX Red
rpOXRpO — Vermilion
rpOXrpO Orange
RPOXRpO Red
RPOXrpO Red
RpOXRpO — Vermilion
RpOXrpO Vermilion
rPOXRpO Red
rPOXrpO Pink
A curious, and I think significant,
relation will be observed between the sexes in the last two classes;
for, the pink females are twice as numerous as the pink males,
while the reverse holds for the orange-eyed flies.
relation comes up again in the ‘pink-male by orange-female
cross to be described later.
The same
CHROMOSOMES AND SEX-LIMITED INHERITANCE 375
This completes the crosses between red and the other colors.
We may now take up the remaining combinations. It will be
noted that from now on the results are about an exact duplication
of the series just described. The results may be said to be mirror
figures of each other which suggests the fanciful idea that the
combinations of colors, that the tables represent, have some such
stereometric relation.
Vermilion eye by pink eye
When a vermilion-eyed female is crossed with a pink-eyed
male, all the female offspring are red, and all the male offspring
are vermilion. These inbred produce in the second generation
all four classes of both sexes:
Red 9 . 110
Red @... 62
Vermilion @ 104
Re ' ast Red 92 Vermilion @.. . 84
Vermilion 2 by Pink o = eatin of » Pink 9 36
Pink @ a 6
Orange 9 . 36
Orange © 17
A deficiency in the males of every class is noticeable in this cross.
The total of all the females is 286 and of all the males 179, nearly
2to1. The analysis, as shown in the next table, calls of course
for equal numbers.
376 T. H. MORGAN
Vermilion 9 RpOX — RpOX
Pink o& rPOX — rpO
Red 2 RpOXrPOX
Fi Vermilion o& RpOXrpO
; = Red 9 RpOX — rPOX — RPOX — rpOX
Gametes of Fi Vermilion o& RpOX — rpOX — RpO-= — rpO
RpOXRpOX = Vermilion 9?
RpOXrpOX = Vermilion 2
rPOXRpOX = Red @
rPOXrpOX = Pink 9°
RPOXRpOX = Red @
RPOXrpOX = Red 9
rpOXRpOX = Vermilion ?
F, rpOXrpOX = Orange 9
RpOXRpO = Vermilion @
RpOXrpO = Vermilion @
rPOXRpO = Red @
rPOXrpO = Pink @
RPOXRpO = Red &
RPOXrpO = Red &
rpOXRpO = Vermilion #7
rpOXrpO = Orange &
The expectation both for males and females is 3 red, 3 vermilion,
1 pink, 1 orange. The females give approximately this result,
while the males fall below the expectation, especially the pink
and orange males.
The reciprocal cross, vermilion male by pink female, gives all
red offspring. These inbred give for the F, generation red males
and females, vermilion males, pink males and females, orange
males. Here again a case of double sex-limited inheritance
occurs, and of course in the same colors, vermilion and orange, as
before.
Red @.. satiate)
| Red & _ 804
pt Gah A) {Red 9 \_J] Vermilion &@. Oot,
Vermilion &@ by pink ° \Red ¢ / Pink? © 214
| Pink @. 99
Orange 84
CHROMOSOMES AND SEX-LIMITED INHERITANCE 377
The numbers are relatively high in this experiment, and signifi-
cant. The red and the vermilion males taken together give
approximately the same number as the red females. Similarly
the pink and the orange males, taken together, are nearly as
numerous as the pink females. The following table shows what
the expectation is in regard to these numbers:
Pink @ rPOX — rPOX
Vermilion « RpOX — RpO
Red 2 rPOXRpOX
IB Red @ rPOXRpO
Red 9 rPOX — RPOX — rpOX — RpOX
Gametes of Fi Red @ rPOX — RPOX — rpO — RpO
rPOXT POX = Pink @
rPOXRPOX. = Red 9°
RPOXrPOX = Red @
RPOXRPOX = Red 9
rpOXrPOX = Pink 9
rpOXRPOX = Red 9
RpOXrPOX = Red @
F, RpOXRPOX = Red @
rPOXrpO = Pink ¢
rPOXRpO = Red &
RPOXrpO = Red @
RPOXRpO = Red &
rpOXrpO = Orange &
rpOXRpO = Vermilion @
RpOXrpO = Vermilion @
RpOXRpO = Vermilion ~
The preceding analysis shows that there should be three times
as many red females as pink females. There are, in fact, some-
what more than three times as many. The pink males should
be to the red males (or to the vermilion) as 1 to 3. They do not
come up to this ratio but nearly approximate to it. Similarly for
the orange males. In this instance, where the numbers are large,
it is quite apparent that the expected and the realized results
fairly agree. The failure is here again obviously due to a reduc-
tion in the number of the pink and orange classes.
378 T. H. MORGAN
Vermilion eye by orange eye
When a female with vermilion eyes is bred to a male with orange
eyes, all of the offspring are vermilion eyed. These inbred pro-
duce the two grand-parental colors in males and in females.
WermiltonOn mcr =. « 909
ae J Vermilion ? \ J Vermilion ¢..... aire LL
a =
MEeeuIbOn Paya Or aneerc \ Vermilion / Orange mOr Geiser 131
Orangesicie cece: 177
The one point to notice here is the excess of orange males over
orange females that has occurred in all of the preceding cases
where both classes occur. The absence of red and of pink from
the combination is due of course to the absence of pink in both
parents.
Vermilion 9 RpOX — RpOX
Orange @ rpOX — rpOX
Vermilion 9 RpOXrpOX
a Vermilion co’ RpOXrpOX
Vermilion @ RpOX — rpOX
HOE Vermilion o RpOX — rpOX — RpO — rpO
RpOXRpOX |
RpOXrpOX } 8 vermilion 9
rpOXRpOX |
F. rpOXrpOX 1 orange @
RpOXRpO |
RpOXrpO 3 vermilion
rpOXRpO |
rpOXrpO 1 orange of
The analysis calls for 3 vermilion females to 1 orange female.
In fact, 53 times as many vermilion as orange females are found
and the same disproportion, due to deficiency in orange, is found
also in the male classes.
CHROMOSOMES AND SEX-LIMITED INHERITANCE 379
The reciprocal cross, vermilion male by orange female, gives
also all vermilion offspring. These inbred give both original
classes in both sexes.
{ Vermilion 9.....411
a { Vermilion Q Vermilion <. 330
7 by Ors =h is a
Wermslionsc) by, Orange. ¢ \ Vermilion Orange ..50
, Orange: ........ 62
Here a slight excess of orange males over orange females occurs.
The relation of vermilion to orange is seen in the following analy-
sis.
Orange 9 rpOX — rpOX
Vermilion « RpOX — RpO
P Vermilion 9 rpOXRpOX
: Vermilion &@ rpOXRpO
Vermilion 9 rpOX — RpOX
Gametes of Fi Vermilion o& rpOX — RpOX — rpO — RpO
rpOXrpOX. =1 orange @
rpOXRpOX
RpOXrpOX ; = 3 vermilion 9
RpOXRpOX
Fy
rpOXrpO =1 orange
rpOXRpO |
RpOXrpO }=3 vermilion <
RpOXRpO J
The expectation is for three vermilion females to one orange
female: the actual numbers are 8 to 1. The same expectation
holds for the male while the actual numbers give 53 to 1.
380 T. H. MORGAN
Pink eye by orange eye
When a pink-eyed female is paired with an orange-eyed male,
all of the offspring are pink. These inbred produce pink males
and females and orange males.
ee Pinks Oneee tee. 1035
Pink 9 by Orange @ = eee °S| Bink he eee eee heb ID
Caste \Orangeug's) seven ser se 518
The two classes of males taken together give almost exactly the
same number as the females. Both pink and orange males occur
in equal numbers. In this experiment there appeared 15 orange
females (not given in the table). This is possibly due to further
mutation in the hybrid or to some error. The formulae are as
follows:
Pink @ rPOX — rPOX
Orange & rpOX — rpO
F Pink @ rPOXrpOX
Pink @ rPOXrpO
G sof F Pink @ rPOX — rpOX
Sate Les souel Pink @ rPOX — rpO
rPOXrPOX Pink 9
rpOXrPOX Pink @
a)
F, rPOXrpO Pink o
rpOXrpO Orange ¢
The reciprocal cross, pink males by orange females, gives the
criss-cross Inheritance, viz., pink females and orange males. It
is of interest to note in passing that there were 233 pink females
and 215 orange males recorded in this F; generation, showing that
in the direct cross the sexes of opposite colors appear in nearly
equal numbers. In the second generation both colors in males
and females occur.
CHROMOSOMES AND SEX-LIMITED INHERITANCE 381
{ Pink 9 572
: J Pinke ) Pink @. 292
Pink & by Orange 9 = Oradeec. |LOrance.e 319
| Orange o.. syarate telnet
In the F, generation a peculiar relation becomes apparent: there
are one half as many pink males as pink females, while in the
orange class the females are only a little more than half as numer-
ous asthe males. Thus while the total number of females of both
classes (884) is nearly the same as the total number of the males
when both classes are added together (814), yet this approximate
equality is due to the reverse ratio of the sexes in the two color
classes. It is worth noting, too, that this relation exists in the
same group, in which as stated above, the pink females and
orange males existed in equal numbers in the F; generation. In
fact, it is just these two classes that still exist in equal numbers in
the F, generation that give the significance to these results.
The analysis of this case is as follows:
Orange @ rpOX — rpOX
Pink & rPOX — rpO
Fr Pink @ rpOXrPOX
Z Orange @ rpOXrpO
: Pink 9 rpOX — rPOX
Gametes of F; Orange @ rpOX — rpO
rpOXrpOX Orange @
P rPOXrpOX Pink @
e ; rpOXrpO Orange
rPOXrpO Pink @
At present I can offer no reasonable explanation of this peculiar
relation between color and sex as shown in this experiment. It
appears to be related to facts to be described later in connection
with associative inheritance, and it is probably also related to a
change in the sex-ratio that I have recorded in another cross
(see Proe. Soe. Exp. Biol. and Med., 1911). As I have these
problems still under investigation I shall not discuss them further
here.
382 T. H. MORGAN
Discussion of results on eye-color
An examination of the formulae used to interpret the preceding
results will show three points of importance. First, that the red
factor R may be present in the male-producing sperm. It is
present there, in fact, if the fly has either red or vermilion eyes.
Second, the factor for pink is only present when X or the sex
factor is present. It is absent, therefore, from all male-producing
sperm. It is true that X may exist without the pink factor, as in
the vermilion and orange flies that owe their peculiarity to the ab-
sence of P. If the P is contained in X, as its connection with
sex establishes, then its absence must be due to its loss from X.
Consequently while X may exist without P, the latter, P, can
occur only when X is present.
Third, the factor for orange, O, is present in every case. It
might, therefore, be omitted from all of the formulae without
affecting the results, provided the absence of R and of P be
assumed to give O. But since orange is a definite color the
absence of red and pink can not be assumed to leave orange.
For this reason I have always inserted it. Its location can not
be identified because it seems never to be lost. I shall give my
reasons later for not identifying it with the color-producer C.
The facts here recorded for the factor P amount in my opinion
to a demonstration that this factor is intimately associated with
the factor for sex. All of the 58 classes found in the second filial
generation can be accounted for on the assumption that X contains
P, when P is present; and, as I pointed out in connection with the
heredity of white eyes versus red eyes, sex-limited inheritance
ean be explained by assuming that X carries red if red is pres-
ent. In the ease of vermilion and of orange eyes pink is lost from
X and the formulae give the classes realized. The asymmetrical
distribution of pink follows the same law as the asymmetrical dis-
tribution of the sex chromosomes.
A study of the formulae also reveals the fact that in the male of
these classes (red and pink) when pink is present in the simplex
condition (it can not occur otherwise in the male) no interchange
takes place between the pink element contained in the X-chromo-
CHROMOSOMES AND SEX-LIMITED INHERITANCE 383
some and any other chromosome, because, as I have previously
pointed out, the sex chromosome in the male has no mate. Con-
sequently no such interchange of chromatic particles, as we must
assume possible for the other chromosomes, is here possible. The
entire scheme of sex-limited inheritance rests, as I have tried to
show, on this simple basis. To prevent a possible misunderstand-
ing I may point out that the behaviour of the R-factor illustrates
how an interchange of R and no Ris possible in the male. If the
R is contained in some other chromosome in the heterozygote it
may interehange position with its absent condition in a corre-
sponding position or particle in the mate of this chromosome.
The facts here recorded for the inheritance of pink make out a
strong case in favor of the view that sex-limited inheritance
can be explained if we locate the factor for pink in one of the sex
chromosomes. I have pointed out that a similar assumption
explains the heredity of white eyes, also sex-limited. I can state
that the same assumption will account for the inheritance of yellow
color and for the two wing mutations that are sex-limited. More
important still is the fact that the extremely complicated results
that follow when two or more of these sex-limited characters are
combined must also be explained on the same principle. It is
this evidence that has convinced me that segregation, the key
note to all Mendelian phenomena, is to be found in the separation,
during the maturation of the egg and sperm, of material bodies
(chemical substances) contained in the chromosomes.
This conclusion need not mean that the material bodies pres-
ent in the chromosomes are the substances out of which the
unit-characters are built up. On the contrary all that this evi-
dence goes to show is that the bodies represent some material
necessary for the development of the particular character in ques-
tion, and it is certain that other parts of the cell also contribute
to the elaboration of the unit-character. This is the view I
should adopt, provisionally, as the more probable. We see, in
fact, that the red color of the eye of the wild fly is due to the col-
laboration of at least four different factors in the cell, namely, a
red, a pink, and an orange determiner, and a color producer.
The pink determiner and the color producer are carried by the
384 T. H. MORGAN
X-chromosome, but are not otherwise related. The red factor
is contained in some other chromosome; the orange factor we
can not yet locate.
Concerning the chemical nature of the three colors I have no
facts to offer. That they may be related chemically is made
probable by the evidence, to be given later, that they are all three
activated by the same color-producer C. If this is admitted we
see that similar substances may be contained in different chromo-
somes, and the further conclusion is then near at hand that in
some cases the same substance may be carried by more than one
chromosome. In connection with a mode of inheritance that is
not as yet clearly Mendelian, viz., beaded and truncated wings
I shall examine this assumption further, but it is not needed for
the cases here described that follow Mendel’s law for one pair of
factors.
In the application of the Mendelian formulae to the F, genera-
tion it has been only too often apparent that while the formulae
give in all cases the expected classes, yet the numerical results
depart widely from expectation. A consideration of the facts
will bring conviction to anyone, I think, that the numerical depar-
tures from expectation are due to special, disturbing factors. At
another time, when I am able to present other data for wing inher-
itance, and for disturbances in the sex ratios, I shall take up the
question of these irregularities more fully. Here I can only
point out one or two possibilities. In some cases the disturbance
can be traced directly to the principle of ‘association.’ By this
I mean that during segregation certain factors are more likely to
remain together than to separate, not because of any attraction
between them, but because they lie near together in the chromo-
somes, as will be explained more fully later. For example, when
red 2 RPOX is crossed with pink # rPO the offspring are red ¢
RPOXrPOX and red « RPoXrpO. As shown by the analysis
on page 370 there are four classes of spermatozoa possible, but if
R and P tend to hold together? rather than interchange with r
there will be more female-producing sperm RPOX than rPOX,
2 In reality C and P.
CHROMOSOMES AND SEX-LIMITED INHERITANCE 385
and hence proportionately more reds in relation to pink than
random or Mendelian segregation demands. The same principle
applied to other cases will often account for the disturbances in
the Mendelian ratios.
It is evident, however, that other conditions also may be respon-
sible for the irregular ratios. The fact that the low types of
mutants—those that have lost two factors, for example—fall
short of expectation as compared with the normal type, and the
disturbances in the sex ratios call perhaps for a different explana-
tion. In regard to the former it is probable either that gametes
containing certain combinations are less likely to fertilize or be
fertilized or that the product of such fertilization is less viable.
Until certain work is completed that I have on hand there is no
need to attempt to decide which of the suppositions is the correct
one.
386 T. H. MORGAN
PARTE ai
THE RELATION OF THE COLOR PRODUCER C TO THE COLOR
DETERMINERS OF EYE COLOR
In a preceding paper (Science, July 1910) dealing with sex-
limited inheritance of white eyes I have shown how the results
are explicable on the assumption that the factor for red color
is absent from the X-chromosome in the white-eyed individuals.
It may appear that this assumption flatly contradicts the assump-
tion made in the preceding cases for eye-color in which it is shown
that the factor (a factor!) for red R is present in the male-produc-
ing sperm (when red is present at all). There is no contradiction,
however, for it is not the color determiner R that is present in
the X-chromosome, but the color-producer C. It is the absence
of C from X that gives the white-eyed fly whose formula is cRPO.
For the sake of simplicity I have not introduced this relation in
the preceding examples. The change, if introduced, gives pre-
cisely the same results, but adds another letter.
I purpose now to consider the relation of this C factor to the
eye colors. It may make the case simpler if first an example
for white and red is given.
When a white-eyed male is crossed to a red female the offspring
are red. These inbred give red females (50 per cent), red males
(25 per cent), and white males (25 per cent). The formulae are
as follows:
Red @ CRX — CRX
White o cRX — cR
Red 9 CRX = cRX
Pi Red 7 CRX — eR
CRXCRX = Red 9
2 CRXcRX = Red 9
: GRKcR| = Red @
CRXeR) = White a
The converse cross, white female by red male, gives red females
and white males. These inbred give red males and females and
CHROMOSOMES AND SEX-LIMITED INHERITANCE 387
white males and females. The formulae for this case are as
follows:
White 9 cRX — cRX
Red o& CRX — cR
Red 9 cRX — CRX
F White @ cRX —cR
eRXcRX = White 9
F CRXcRX = Red 9
ecRXcR = White @
CRXcR = Redo
These two examples will serve to show the method by which all
the white-red combinations can be treated. The results are those
that I have already published (Science, 1910, vol. 32).
Pink eye by white eye
The results of this combination have been already given
(Science, 1911). Since the numerical relations were peculiar
I repeated the experiment and obtained large numbers of individ-
uals that furnish a better basis for interpretation. When a
pink-eyed female is bred to a white-eyed male all the offspring
have red eyes. These inbred produce red-, white-, and pink-
eyed offspring in the following proportions:
Red-eyed females............ ro A oes 0 ce ENCI 1183
Red-eyed males............
White-eyed males.............
Pink-eyed females.........
Pink-eyed males...........
If we interpret these results in the same terms as those used
for white and red we get the following formulae. Since the orange
factor is present throughout and the orange eye is not involved
O is omitted. The white-eyed male came from red stock through
the loss of C and his formulae is cRP + cR.
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. I1, No. 4
a4
388 T. H. MORGAN
Pink 9 (CrPX: — iCrPxX
White «@ cRPX — cR
ted 9 CrPXcRPX
Red & CrPXcR
Gatcciesoe an Red @ CrPX — crPX — cRPX — CRPX
+ e : Red @ CrPX — CRPX — cR — cr
CrPXCrPX = Pink @
CrPXerPX = Pink 9?
CrPXcRPX = Red @
CrPXCRPX = Red @
CRPXCrPX = Red 9
CRPXerPX = Red @
CREXcRPX | 3= Red @
CRPXCRPX = Red 9
F, cRCrPX = Red &
ceRerPX = White @
eReRPX = White @
cRCRPX = Red &
erCrPX = Pink ¢
ercrPX = White @
ereRPX = Whites
erCRPX = Red &
The expectation is six red females to two pink females or 3 to 1.
The realization is close to expectation, there being a small deficit
in pink females. For the males the expectation is three reds to
one pink. The realization is 4 to 1 due to deficit in pink males.
The expectation for white males is the sum of the pink and red
males: the realization is not far from this number. The white
males should be to the red males as 4:3 which is approximately
realized.
The reciprocal cross, white-eyed females to pink-eyed males,
gives in the first generation red-eyed females and white-eyed
males. These inbred gave the following colors and ratios:
CHROMOSOMES AND SEX-LIMITED INHERITANCE
Red-eyed
Red-eyed
White-eyed
White-eyed
Pink-eyed
Pink-eyed
389
ORME MOC PETERS py Aloo. 900, soy ie Wc Vee AG 706
CRE ne ae tar ERA erent 747
ORM reel eRe eat Fr elon ae er) eae S04
Fo enchants oy oe Re CT eae, A RPTL OP ear Ptr eS ae 832
OREN al hats ON ee ee en 8 Oe ee ee 204
Efe PR er: cere hs entree Ae Puree tated ls 210
the same formulae to this case gives the
The application of
following results:
F
Gametes of F;
Red
White
White. 9 cRPX — cRPX
Pink @ CrPX — cr
Red @ cRPXCrPX
White o« cRPXcr
@ eRPX — CrPX —
o& cRPX — crPX — cr
cRPXcRPX = White 2
cRPXerPX = White 9?
CrPXcRPX = Red @
CrPXerPX = Pink ?
CRPXcRPX = Red @
CRPXerPX = Red 92
erPXcRPX = White 9
erPXerPX = White 9?
ceRPXer White 7
cRPXcR White @
CrPXer Pink @
CrPXcR Red &
CRPXer Red ¢&
CRPXcR Red ¢&
erPXer White @
erPXceR White @
In both males and females the color ratio
The actual numbers are a fair approximation to this expectation.
CRPX — crPX
is 4 white, 3 red, 1 pink.
390 T. H. MORGAN
PART III
HEREDITY OF TWO SEX-LIMITED CHARACTERS COMBINED
WITH FOUR EYE COLOR CHARACTERS.’
In this experiment a male with short proportionate wings and
white eyes—both sex-limited characters—was mated toan orange-
eyed female with long wings. The white-eyed male was a white
from red stock, cRPO.
In the first generation all of the offspring had long wings like
the mother’s; the females had red eyes and the males had ver-
milion eyes. These were inbred and produced the second or
F, generation that contained flies having red, vermilion, pink,
orange, and white eyes. In each of these classes, however, the
short-winged individuals were males, as shown in the next table:
Red @ long wings eS eee so pega roe eget ae Ou
Red) ict long; wintss-ceessescre aster sets sn, VERE eee See 14
Redo short. win scree ae eee PPR Ee ec seed 6 2 66
Vermilion 92 long wings.. Sea SaTKe t Aaa MG hunts ate)
Mermilion "67! Long) Wiles x. ses, eee eds cceeaneds verbal ore teats .. 204
Vermilion & short wings Bc SME RIE See peonae il)
Pink @ long wings oe ; =e .. 88
Pink o& long wings ; 5 : Se oes 7
Pink @ short wings : ; , Ci ee.
Orange 2 long wings.. as Sethe Bh enc 119
Orange @ long wings.... ba fase eS : : erigedi)
Orange o& short wings........... By creates ; mols eal
White & long wings eGo: Wathadee 95
White o short wings tops Trea : 232 2202
It will be observed at once that the inheritance of short wings
and of white eyes is strictly sex limited. It will also be observed
that each eye color has been combined with short wings, but
only of course in the male sex. The total number of short winged
males having red, vermilion, pink, and orange eyes is 92, while
3 See also ‘“‘The Method of Inheritance of Two Sex-Limited Characters in the
Same Animal.’’ Proce. Soc. Exp. Biol. and Med. vol. 8, Oct. 1910.
CHROMOSOMES AND SEX-LIMITED - INHERITANCE 391
the number of short winged males with white eyes is twice this
number. In this connection I wish to point out that the grand-
father had white eyes. The possible significance of this may be
discussed later. It is obvious, however, on any theory of chance
elimination- of unit characters in the egg that the total number
of males of all eye colors having short wings must be equal to
the number of short winged white eyed males.
The analysis of the results following the same methods as here-
tofore is as follows:
Long-winged, orange-eyed 9 LCrpOX — LCrpOX
Short-winged, white-eyed o IeRPOXIcRpO
Long-winged, red-eyed 9 LCrpOXIlcRPOX ; Fr
Long-winged, vermilion-eyed « LCrpOXIeRpO L
LCrpOX LerpOX
Py LOR pO , |LeRpOX
3 |LCrPOX $ | LerPOX
2 = |LCRPOX = |LeRPOX
CERCHES I % )1CrpoX & | lerpOX
&) |ICRpOX © |1ecRpOX
= ICrPOX | lerPOX
ICRPOX leRPOX
é g 8 {LCrpOX E FI a) LCRpOX
& §, © \lcRpO Bo ® \ lerpO
The random fertilization of these sixteen kinds of eggs by the
two kinds of female-producing spermatozoa LCrpOX and LCR-
pOX and their fertilization by the two kinds of male producing
spermatozoa IeRpO and lerpO is represented in the following
table:
392
Egg
1CrpOXLCrpOX
ICRpOXLCrpOX
ICrPOXLCrpOX
ICRPOXLCrpOX
LerpOXLCrpOX
LerpOXLCrpOX
LerpOXLCrpOX
LeRPOXLCrpOX
lerpOX LCrpOX
leRpOXLCrpOX
lerPOXLCrpOX
IeRPOXLCrpOX
Egg
LCrpOX1CrpO
LCRpOX1CrpO
LCrPOXICrPO
LCRPOXI1CrpO
ICrpOX1CrpO
ICRpOX1CrpO
1CrPOX1CrpO
ICRPOX1CrpO
LerpOX1CrpO
LeRpOX1CrpO
LerPOX1CrpO
LeRPOXI1CrpO
ICrpOX1CrpO
ICRpOXICrpO
1CrPOX1CrpO
ICRPOXICrpO
Sperm
LCrpOXLCrpOX
LCRpOXLCrpOX
LCrPOXLCrpOX
LCRPOXLCrpOX
Sperm no
long orange
long vermilion
long pink
long red
long orange
long vermilion
long pink
long red
long orange
long vermilion
long pink
long red
long orange
long vermilion
long pink
long red
long orange
long vermilion
long pink
long red
short orange
short vermilion
short pink
short red
long orange
long vermilion
long pink
long red
short orange
short vermilion
short pink
short red
40 10 40 10 40 40 40 140 40 40 410 +40
¥™™A™A 40 40 10 40
AAqQyA
WA A™®
AAA™A
and
and
and
and
and
and
and
and
» MORGAN
Sperm X
LCRpOX
LCRpOX
LCRpOX
LCRpOX
LCRpOX =
LCRpOX
LCRpOX
LCRpOX
LCRpOX
LCRpOX
LCRpOX
LCRpOX =
LCRpOX
LCRpOX
LCRpOX
LCRpOX
long vermilion
long vermilion
long red
long red
long vermilion
long vermilion
long red
long red
long vermilion
long vermilion
long red
long red
long vermilion
long vermilion
long red
long red
Sperm no X
IeRpO
leRpO
IeRpO
IeRpO
leRpO
IcRpO
leRpO
IecRpO
IcRpO
IeRpO
IceRpO
IeRpO
leRpO
IeRpO
IecRpO
leRpO
long vermilion
long vermilion
long red
long red
short vermilion
short vermilion
short red
short red
long white
long white
long white
long white
short white
short white
short white
short white
AAAA QAAqaa 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40 40
WA A ®™
WAAA
CHROMOSOMES AND SEX-LIMITED INHERITANCE 393
Summarizing this last table we get:
Long winged red 9...12 Long winged red @ 4
Long winged vermilion 9...12 and Long winged vermilion o7 4
Long winged pink @... 4 Long winged pink @ 2
Long winged orange 92 4 Long winged orange a2
Long winged white ~ 4
Short winged red 1
Short winged vermilion = 4
Short winged pink @ a2
Short winged orange @ ees
Short winged white o..... 1
There should be, obviously, as many females as males. The
results show 1012 females and 7428 males. There is a distinct
falling off of males. The two sexes give about 10 to 8 or 5 to 3.
Tn each class with eye color the males should be the same in
number in the red and the vermilion; and in the pink and the
orange. A great variability is however realized as shown below:
RED VERMILION PINK ORANGE | WHITE
One MwA eC ae sciep-r)- tyre ms 14 254 7 79 95
Short winged @ 66 10 15 1 202
The enormous discrepancy between theory and fact shown by the
table may well make one reject the theory as totally inadequate
to explain the facts. Nor can one appeal to the relative viability
of the males to help him out of the dilemma; for, the long winged
normal males run far behind the number for the vermilion, orange,
and white; yet the long winged red eyed males are normal for the
species, and do not run behind under the same conditions used in
this experiment. A closer scrutiny of the table will, however, indi-
cate a relation that may be very significant. The great excess of
males is found in two classes, the long winged vermilion and the short
winged white male, and these are respectively the father (the same
as the grandmother’s combination) and the grandfather of this PF,
generation!
394 T. H. MORGAN
PART IV
THE INHERITANCE OF THREE SEX-LIMITED CHARACTERS
In the following four crosses three sex-limited characters are
involved: white eyes, short wings; and yellow color. In the first
two crosses all these characters are contained in one of the par-
ents; in the other two crosses two of the characters are in one
parent and one in the other.
In earlier papers (Science, 1910 and 1911) I have described
the main facts for inheritance of red versus white eyes, and long
versus short (short proportionate) wings. In order to understand
the relation of yellow to normal color, the following experiment
may be cited. A long, red, normal (body color) female was
crossed with along, red, yellow male. The offspring, both male
(654) and female (705), were long, red, normal. These inbred
gave:
Normal @ Se cH favs Sunken ray fereyativiasct toe OP MMSE SR ASHES AT Cue TRIN: SETS 525
Normal o.. BRR e MTG paler ats iti ciaehe Be Se, CEE yee a TE 340
Yellow ¢& Lee id RO CTO OIE Ceti Sree eae Oe beac 194
The reciprocal cross, viz., long, red, yellow female by long, red,
normal male, gave females (397) with long wings, red eyes, nor-
mal color and males (282) with long wings, red eyes, and yellow
color. These inbred gave:
S07 117-1 BRS Pe One arr SS I dR SA oa Ln nk A mam ie, As a 0 346
Wormiall Gus ste he eee irs So ee ee ee cya ce Oe Sk ee 259
MEU ons: Botton ole RETRO eron Ca aah CAL A cua ee aA REI ERIN ie Ge ea 226
VSN » ohh Se oi ae NtS cinch ERAT I TEE SPARRO Aa mre 230
We are now in position to take up the experiments in which three
sex-limited characters, white eyes, short wings, and yellow color
are involved; see Plate I.
When a yellow, short winged, white eyed male is bred to a
normal wild fly with normal color, long wings, and red eyes, all
the offspring are like the mother, i.e., normal color, long wings,
red eyes.
CHROMOSOMES AND SEX-LIMITED INHERITANCE 395
aoe, ‘ _ JNRL 9? = Normal red long 9
VWs cp by NRL e — (NRL o& = Normal red long #
These inbred have produced four main classes of males, and
three small classes, represented by a few male flies only; the females
are represented by but a single class, as shown below:
Normal. Color
RED EYES | WHITE EYES
Long wings | Short wings Long wings Short wings
g | fo} | ) fof g of g a
1879 | 606 | 167 1 3
| |
Yellow Color
RED EYES WHITE EYES
Long wings Short wiugs Long wings Short wings
= = —- =
Q | fol 9 (ofl 2 fol g fol
7 1438 | 96
i
In addition to the flies recorded in the table there were two females
belonging to two classes, viz., one, normal, long, white female;
and one, yellow, long, white female. I shall not hesitate to ignore
these two cases as exceptional, due either to accidental contam-
ination through the food, or to sporting within the stock. Omit-
ting these two females it will be seen that all of the females fall
into one class having normal color, long wings, and red eyes. In
all there were 1879 of these females, as against 1022 males.
The females are, therefore, almost twice as numerous as the males.
The three sex-limited characters appear only in the grandsons;
one class containing all three sex-limited characters, yellow,
short, white (96); one class containing two sex-limited characters,
yellow, white (143); and one class containing one sex-limited
character, short wings (167). For the moment the other three
classes of males may be left out of account. It will be noticed
also that the grandfather’s combination is well represented by 96
396 T. H. MORGAN
individuals, while the father’s combination which is the grand-
mother’s also is represented by the great majority of all the males
(606). The remaining large class contains two of the grandfather’s
characters, viz., yellow and short. Of the three small classes of
males, one contains two of the grandfather’s characters, viz.,
white and short, and one contains one of his characters, viz.,
yellow color.
In my first attempt to analyze this case I ignored the three small
classes of males because I found empirically that the remaining
classes and the females could be very simply accounted for, as
the following formule will show.
Normal, red long @ NRLX — NRLX
Yellow, white short o YWSX — —
Normal red long @ NRLXYWSX
uy Normal red long o NRLX ——
NRLX — NRSX — YWLX — YWSX
Gametes of F; NRLX —
F, Generation
NRLXNRLX = Normal red long 9 — NRLX = Normal red. long rou
NRLXNRSX = Normal red long 9 — NRSX = Normal red short fou
NRLXYWLX = Normal red long 9 — YWLX = Yellow white long fou
NRLXYWSX = Normal red long 9? — YWSX = Yellow white short @
This scheme meets with two serious difficulties. It calls for
equal numbers of each kind of male while in reality one class
is at least three times as numerous as any one of the others. If
we tried to explain this anomaly (as in fact I think we must)
on the basis of some sort of “association”? taking place, we still
have to meet a more serious theoretical difficulty. It will be
seen that only four classes of eggs are represented in the F, gen-
eration. There are two classes of eggs containing N and R,
and two containing Y and W, but no class containing N and W,
and none containing Y and R. No theoretical explanation can
admit this arbitrary treatment, for the theory on which we are
working demands the full interchange of all of these characters—
unless some special reason can be given for failing in this regard.
CHROMOSOMES AND SEX-LIMITED INHERITANCE 397
Now it will be seen that the combinations of N and R, and Y and
W are the combinations that existed in the parents of this cross.
To admit the foregoing scheme requires the recognition of this
union as permanent in subsequent generations. Yet this is op-
posed to the Mendelian treatment of the case, unless association
is admitted as valid.
If now we take up the case of the three small classes of males
we find no place for them in this scheme. I see no reason for
ignoring them, small though the classes be. This consideration
leads me to the conclusion that instead of four classes of eggs
in the F, generation, the possibility of eight classes must be
admitted, but owing to the initial association of N and R, and Y
and W, their separation only occasionally occurs. When it does,
the small classes of males appear, and the number of individuals
in these classes is a measure of the infrequeney with which the
separation occurs. The scheme when fully worked out is as
follows:
Normal female NRLX — NRLX
yellow white short « YWSX — ——
F Normal 9 NRLXYWSX
: Normal o& NRLX —
Gametes of F;
NRLX — NRSX — NWLX — NWSX — YWSX — YWLX — YRSX — YRLX
NRLX = ——
NRLXNRLX = Normal red long 2 > — NRLX = Normal red long ou
NRLXNRSX = Normal redlong 9 — NRSX = Normal red short of
NRLXNWLX = Normal red long 9 — NWLX = Normal white long @&
NRLXNWSX = Normal red long 9 — NWSX = Normal white short &
NRLXYWSX = Normal red long 2 — YWSX = Yellow white short rot
NRLXYWLX = Normal red long 2 — YWLX = Yellow white long of
NRLXYRSX = Normalredlong 9 — YRSX = Yellow red short fot
YRLX = Yellow red long of
NRLXYRLX = Normal red long ?
It is seen that the two factors N and W (or Re) tend to hold to-
gether. Both are contained in the sex chromosome, i.e., N and
e are there. Both are present in the grandmother, and through
398 T. H. MORGAN
her carried into her son—the father of the F, generation. The
grandmother transmits only one X to her grandson—the one in
question. It is also seen that the two factors Y and R tend to
hold together in the same way. On the other hand the factor
for long and short wings seems freer to leave one X and pass to
its partner without showing any very great tendency to associate
with the color factors in X.
The reciprocal cross, viz., a female with yellow color, white
eyes, and short wings, bred to a normal male with normal color,
red eyes, and long wings, gave females with normal color, red
eyes, and long wings, and males with yellow color, white eyes,
and short wings.
J NRL ¢ = normal red long Q
SRS ees A | YWS o& = yellow white short ~
These inbred gave the classes in the next tables:
Normal Color
RED EYES WHITE EYES
(=e ; a. Sa Se oT,
Long wings Short wings Long wings | Short wings
9 a | 9g o On Ollaac OF aaecr
|
439 319 | 208 193 1 5 11
Yellow Color
RED EYES WHITE EYES
——_ = i
Long wings Short wings Long wings Short wings
= | | >| wh es
Q Si ane st || 2 foil g fof
Here again the two pairs of grandparental characters, viz., nor-
mal color with red eyes; and yellow with white eyes, are repre-
sented by the eight large classes in the F, generation; while short
and long wings are nearly equally distributed. But even here
there are more grandchildren with normal color, red eyes, and
CHROMOSOMES AND SEX-LIMITED INHERITANCE 399
long wings, than with short wings. ‘These two colors went to-
gether with long wings in the grandfather. Conversely, the
grandmother combined short wings with yellow color and white
eyes, and there is an excess of short winged grandchildren ( 2
and ¢) over long winged (and ¢). The shorter analysis is as
follows:
Yellow, white, short @ YWSX — YWSX
Normal, red, long o& NRLX — ——
Normal @ YWSXNRLX
Bi Yellow, white, short « YWSX ——
5 YWSX — YWLX — NRLX — NRSX
Gametes of F; SANDS =
YWSXYWSX = Yellow white short 9 — YWSX_ Yellow white short
YWSXYWLX = Yellow white long 9 — YWLX Yellow white long
YWSXNRLX = Normal red long 9 — NRLX Normal red _ long
YWSXNRSX = Normal red short 9 — NRSX Normal red — short
EQ Qyiay a,
If the more extended analysis for the gametes of the female
were used there would be four more classes of eggs, namely,
YRSX, YRLX, NWLX, NWSX, which would give four new
classes of females, namely, yellow, red, short; yellow, red, long;
normal, white, long; and normal, white, short; of which the
second, third, and fourth are represented in the table by seven,
one and four females respectively. The extended analysis would
also give four other classes of males, whose formulae correspond
to those of the four new types of eggs given above in the text, of
which two are realized and two are not.
In the third cross, a female with normal color, white eyes, and
short wings was bred to a male with yellow color, red eyes, and
long wings. The female offspring had normal color, red eyes,
and long wings, and the male offspring had normal color, white
eyes, and short wings.
es : is {NRL @ = Normal red long 9?
NWS 9 by YRL g' = \NWS @ Normal white short o
400 Wie sie
MORGAN
The F,’s inbred gave the classes shown in the next table:
Normal Color
RED EYES WHITE EYES
Long wings | Short wings Long wings Short’ wings
g | rot g ou 2 fof | rofl
439 | 7 235 218 237 359 | 387
|
Yellow Color
RED EYES WHITE EYES
Long wings Short wings Long wings Short wings
ees = - “ ed: SS
2 fot ie) fol g ou ol ofl
345, 210 | 4
in the second generation there are four kinds of females and six
kinds of males. If we recognize the union of N and W and Y
and R in the gametes of F, (the union that was present in the
grandparents) the expectation on the shorter analysis is as follows:
NWSX — NWSX
YRL — —
Normal, red long 9 NWSXYRLX
Bi Normal, white short «@ NWSX ——
YRLX — YRSX
Gametes of F, NWLX — Rath
NWSXNWLX = Normal, white long 2 — NWLX = Normal white long rot
NWSXNWSX = Normal, white short 9 — NWSX = Normal white short ¢&
NWSXYRLX = Normal red long 9 — YRLX = Yellow, red, long ree
NWSXYRSX = Normal, red short 9 — YRSX = Yellow, red short oft
If we admit the more extended segregation, the same dispropor-
tions appear, but the two smaller classes of males are now repre-
sented and two classes of males do not appear at all (as in the
CHROMOSOMES AND SEX-LIMITED INHERITANCE 401
realization). In regard to the couplings in these cases it is of
great importance to notice that once more the long and short
factors segregate without regard to the color factors, yet even
here a remarkable fact comes to light. The two classes of males
that exceed the others are those in which the long, red, yellow
combination and the short, white, normal combination exist.
These are the two combinations that were in the grandparents.
If we assume that they more often remain in the same chromo-
some the numerical results become apparent.
In the fourth cross, a male with normal color, white eyes, and
short wings was bred to a female with yellow color, red eyes, and
long wings. The female offspring had normal color, red eyes,
and long wings, and the males had yellow color, red eyes, and
long wings.
=) NRL @ = Normal red long
INNIS) eh 3 OS YRL & = Yellow red long
The second or F, generation is represented in the next table:
Normal Color
RED EYES WHITE EYES
Long wings Short wings Long wings Short wings
= = | = =
g fof S| GE i ee isi g | fos!
608 2 137 | i oho37
Yellow Color
RED EYES WHITE EYES
Long wings | Short wings Long wings Short wings
2 fof Q cof gl X2) of g of
389 | 248 101 1 1
Only four large classes of males are represented and it is signifi-
cant that these are the yellow-red and the normal-white—the
two combinations that correspond to the two grandparental com-
402 T. H. MORGAN
binations. These two classes oceur both in the long and short
wings, but here again occurs the significant fact that the yellow,
red, long are twice as frequent as the yellow, red, short. The
former is the grandmaternal combination. Again the normal,
white, short are nearly twice as numerous as the normal, white,
long and it is the former combination that is characteristic of
the grandfather. The shorter analysis follows:
YRLX—YRLX
NWsSxX — ——
F Normal red long 9 YRLXNWSX
: Yellow red long @ YRLX —
YRLX — YRSX — NWLX — NWSX
Gametes of F, En
YRLXYRLX = Yellow, red,long 9 — YRLX = Yellow,red,long . &
YRLXYRSX . = Yellow, red, long @ — YRSX = Yellow, red, short fou
YRLXNWLX = Normal red, long 9 — NWLX = Normal, red, long ol
YRLXNWSX = Normal red, long @ — NWSX = Normal, red, short fe
The same objections may be urged against this scheme that have
been given for the first short scheme. It is unnecessary to write
out the longer scheme again in this case as the same principle
employed in the first instance is applicable here. Of the four
additional classes of males called for by the longer analysis,
three appear in the results represented by any 2, 1, and 1 males
respectively.
CHROMOSOMES AND SEX-LIMITED INHERITANCE 403
PART V
CONCLUSIONS
A THEORY TO ACCOUNT FOR ‘ ASSOCIATIVE” INHERITANCE
In the preceding pages I have tried to show how the mechanism
that exists in the chromosomes ean be applied to the mechanism of
heredity, provided we deal with particles or chemical substances
in the chromosomes rather than with the chromosomes as units.
The evidence makes out, I believe, a very strong case in favor of
the idea that sex-limited inheritance is.connected with the same
physical body that determines sex, and I have not hesitated to
identify that body with the sex chromosome. The second point
of significance in the results is that while in the female there may
be an interchange between homologous chromosomes, no inter-
change takes place in the male of those factors connected with
sex-limited inheritance. We can explain this result if these char-
acters are contained in the single X chromosome in the male
which alone has no mate. The third point of interest in these
results is the necessity of assuming some combination or rather
localization amongst some of the substances resident in the same
chromosome. The peculiar ratios found in the second genera-
tion find their explanation only by means of such an assumption.
Couplings and linkages have been described before to account
for observed ratios, notably by Bateson and his collaborators, but
I think we see here clearly for the first time that these unions are
not due to inherent relations, or fusions, or attractions, or corre-
lations, or repulsions, but to juxtaposition of particles in the chro-
mosomes. It has been shown in a considerable number of cases
that at one stage in the process of union of homologous chromo-
somes the members of each pair twist around each other like
the components of a rope. Subsequently these twisted chro-
mosomes fuse together and shorten. Later a longitudinal split
appears in the shortened, double chromosome. This split now
lies in one plane, i.e., it does not follow the turns of the united
chromosomes. In consequence of the position of the new plane
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, NO. 4
404 ; T. H. MORGAN
of splitting or division each half, or new chromosome, must be
made up of parts of one and parts of the other of the two original
chromosomes that united in pairs as Janssens has shown. As a
result of the subsequent ‘‘reduction”’ division the cells that are
produced will contain new combinations of the materials com-
posing the original chromosomes. If the chromosomal materials
that represent the factors of heredity are placed lineally along
the chromosome and in corresponding linear series in each pair
of homologous chromosomes, random separation of these mate-
rials will be brought about by means of the cell mechanism,
explained above, except in those cases where the materials lie
near together. In the former case, the usual Mendelian random.
segregation will take place; in the latter case, groups of factors
will tend to remain together or be assocfated in heredity. These
latter cases correspond to those in which we find ‘‘association”’
to occur. It will be observed that while such associations will
be more or less common according to the nearness of the associat-
ing factors in the chromosome, the associations are not absolute
for occasionally the twisting of the chromosomes will be such
that even regions lying lineally near together will come to le
on opposite sides of the united chromosomes. These cases repre-
sent the small classes observed in the tabtes. In the case of the
X-chromosomes we should expect interchanges of the postulated
kind to occur when two X’s are present, as in the female of Droso-
phila; but no interchanges when only one X is present as in the
male. The experimental results. accord completely with this
‘anticipation and afford strong evidence in favor of the view ex-
pressed above. This assumption seems to be far simpler than
the assumption of attractions, repulsions and complicated ratios
that Bateson has suggested as an interpretation of similar phe-
nomena.
it is obviously not essential to this hypothesis of factoral inter-
change to limit it to the particular stage of the chiasma type here
suggested, for if a similar phenomenon occurs at any other stage
in division the same results will follow.
.
CHROMOSOMES AND SEX-LIMITED INHERITANCE 405
DISINTEGRATION OF A SPECIES AND ITS RECONSTITUTION BY
RECOMBINATION
The series of mutations that have appeared in Drosophila
can all be accounted for on the assumption of losses from the orig-
inal germ-plasm. There is possibly one exception to this rule,
namely, the melanitic or black mutant that may appear to add
something to the original color. When crossed to the normal it
gives an intermediate type in the first generation, and this fact
also might be urged in favor of the change being in a positive
direction. But since in the second generation the black fly
appears as the recessive type it is not improbable that even this
mutant may be due to loss. The most convincing proof that
the eye and wing mutations are due to loss is found in the recon-
stitution of the original or wild type when certain recombinations
aré made. So many examples of this have been given in the pre-
ceding pages that I need not go over the evidence again. Two
points, however, may be recalled as instructive. In several com-
binations the female alone is reconstituted. In the older term-
inology the female is the atavist, the male is the neomorph;
and it is always the female and never the male that shows this
relation if either one is atavistic. The reason for this is also
clear from the evidence—the male-producing sperm has all
- the recessive facters in question, while the female-producing
sperm has one ol more dominant factors. No better example
than this one of unisexual atavism, or reconstitution, could be
cited to show the advantage that the modern explanation of
heredity has over the older view, where a fact of this kind would
have seemed totally inexplicable.
The second point of interest in this evidence is found in the
different behavior in heredity of the original and of the reconsti-
tuted (atavistic) type, for while the wild type-continues to breed
true the reconstituted type splits later into its components. The
reconstituted type may be said to be physiologically complete,
but morphologically dismembered. For example, when a ver-
milion male is bred to a pink female, all of the offspring are red.
The four substances, R P O C, necessary to produce red have
406 T. H. MORGAN
been brought together, so that all the elements of the wild fly
are present that collectively give red eye-color; but now instead
of the chromosomes that carry these substances being represented
in duplicate, some of the chromosomes lack one or the other
substance. It is due to this that the splitting takes place in the
formation of the germ-cell of these reconstituted types. Never-
theless it is possible even to reconstruct the original wild forms
by suitable combinations, such, for example, as will give in duplex
the four factors essential to the development of red eyes, and the
structurally reconstituted types will breed as true to type as the
wild flies, in contrast to the atavists that are only physiologically
reconstituted.
THE PRESENCE AND ABSENCE THEORY IN RELATION TO THE
. THEORY OF ASSOCIATION
The appearance of so many mutants, due to losses, raises the
question as to whether all new types that follow Mendel’s law may
come under the same category. There are types, it is true, among
domesticated forms that appear to have added something to the
original type from which the mutants arose, but some of these
are due to hybridization, and some may be due to losses of jnhib-
iting factors, whose absence permits the further elaboration of
characters already present. As yet the evidence is insufficient,
I think, to allow any certain generalization in this regard, yet
the evidence suffices at least to show that many or even most
cases that follow Mendel’s law fall under this head. In so far
as the mutants are due to losses they are explicable on the pres-
ence and absence theory, which may seem to give some grounds
for the universal application of this principle. There is, however,
one possibility that we can not afford at present toignore, namely,
the evidence of ‘‘association”’ which is clearly furnished by some
of the crosses described in the preceding pages. I sheuld like
to dwell a little further on this point. If I am right in explaining
the results of those cases where two or three sex-limited charac-
ters are involved on the grounds of the juxtaposition of substances
(factors) in the chromosomes, it 1s only a step further to cases,
CHROMOSOMES AND SEX-LIMITED INHERITANCE 407
like those of the gray mouse when the color factors for gray,
namely, black, yellow, chocolate, and ticking, remain permanently
associated when crossed with a mutant which contains only one
of these same factors. Thus when a yellow bearing germ cell
meets one bearing gray, all of the offspring are yellow. These
yellows inbred produce only grays and yellows and not blacks and
chocolates also, as should happen, did each of the elements in
the original gray have for its mate the absence of its particular
factor.t. In Drosophila the eye color is made up of three or four
color factors. When a wild female is crossed to one of the mutants
an orange eyed male for example, all of the offspring are red.
In the F, generation not only red and orange, but pink and bright
red also appear, although even here there is a stronger tendency
for the original red to reappear more often than its products. It
seems to me probable at least that the difference between the
mice and the fly in this respect may be due to the closer associa-
tion of the factors in the mice than in the fly. If so, the differ-
ence is one of degree only and not of kind.
THE FERTILITY OF DEFICIENT MUTATIONS
A striking falet in regard to most if not all of these mutations in
Drosophila is their infertility compared with the original stock
kept under identical conditions. As I have this matter under
investigation I wish here to touch on it very briefly, and only in
so far as it bears on the numerical proportions of the different
types. The pure stock of several of the new types is less produc-
tive than the original stock, and the failure in several cases of
the deficient types to appear in the expected ratios suggests that
this failure is due in part to the failure in fertility or in vitality
of the new types. Whether this is due to failure in the develop-
ment of the egg, or of the sperm, or to failure to fertilize, or to lack
of development of the embryo, are points requiring special study,
but the facts are sufficiently numerous to raise the question as
to whether a type that lacks some material present in the original
4See Morgan, T. H. The influence of the environment and of heredity on the
inheritance of coat color in mice. New York Academy of Science. 1911
408 T. H. MORGAN
stock may not in many cases lose also its full power of productiv-
ity. It may seem improbable that the presence of some substance
necessary for the development of a particular color in the eve
could have any influenceon the rest of the germ, but the same sub-
stance that is essential for eye color may be essential for the pro-
duction of other things in the body that are not so apparent. In
fact, I have already obtained evidence in the case of the eye
color-producer, C, that the absence of C not only affects the eye-
color, but other parts in the body as well. It is not impossible,
therefore, that in certain cases the absence of a factor may have
an important influence in one or another way on the productivity
of the animal. I do not wish to discuss further this question until
I can bring forward certain evidence that bears directly on it,
but I have raised the question here, first in order to point out its
possible bearing on the disturbance of Mendelian ratios, and
second, in order to make clear that while I regard the evi-
dence in favor of the mosaic inheritance of certain characters as
established, I am not unappreciative of the fact that a simple
factor may have a wider influence in development than appears
when only a single character is under consideration.
ORIGIN OF MUTATIONS THROUGH CHROMATIN LOSSES
The mutations that have occurred in Drosophila may throw
some light on the origin of mutations in general. If my analysis
is correct it follows that mutations arise not through losses of
whole chromosomes as some cytologists have hinted (for the
individual chromosomes must be supposed to earry not one but a
host of factors) northrough the doubling of one or of all of the chro-
mosomes (which would only add to what is already present in
duplex) but mutations arise through the regrouping of partic-
ular substances carried by the chromosomes. These substances
may be so small in amount that their absence may entirely escape
a cytological examination. It would seem that ample opportunity
must be present for losses of this kind, since any irregularity in
the division of the chromatin in the germ tract would lead to
the appearance in time of a mutant if such were viable and if the
CHROMOSOMES AND SEX-LIMITED INHERITANCE 409
right combinations were brought about. If, for instance, in the
division of the spermatogonial cells the material particles at any
level should fail to divide when the rest of the chromosome divides,
one of the resulting cells will be deficient in the substance in ques-
tion, and its offspring will be correspondingly deficient. Or if
after synapsis similar particles in homologous chromosomes should
pass into one chromosome, instead of segregating, one of the result-
ing cells will be deficient. The wonder is that such losses are so
infrequent.
THE BEARING OF THE RESULTS ON THE CONSTITUTION OF THE
SEX CHROMOSOMES
The experiments on Drosophila have shown that a most compli-
cated series of facts relating to sex-limited inheritance can he
accounted for, as I pointed out in my paper of 1910, on the assump-
tion that one of the factors for such characters is combined with
the sex factors, X, or more specifically (Morgan,® Wilson 1911)
if it is contained in the accessory or sex chromosome. The ab-
sence of this chromosome in half of the spermatozoa, and the
impossibility of an interchange between this simple X-chromo-
some in the male (since X has no pair in synapsis) is the signi-
ficant feature of the explanation. What is most important, is
the discovery that the X-chromosome contains not only one of the
essential factors in sex determination, but also all other characters
that are sex-limited in inheritance.
The discovery of this relation leads us a step farther, I think,
in the analysis of the problem of sex determination, for it
shows that the determination of sex is only one of several (per-
haps of a large number of) properties contained in the sex chro-
mosomes. Only by the loss of a factor from one of the other X’s
(in the female) is it possible to aiscover just how many factors
are containedin X. Already four or five such losses have appeared
in my mutations. This leads at once to the inference that it is not
the X-chromosomes, as such, that is a factor in sea determination,
bul only a very small part of its material.
5 In a paper read before the American Society of Naturalists, December 29,
1910. Published in American Naturalist, 1911.
410 T. H. MORGAN
I suggested in 1905 that the female sex is determined by the
presence of more chromatin in the fertilized egg. Wilson sug-
gested in 1906 that it is a particular chromosome that gives the
quantitative results (or at least a more or less ‘active’ chromo-
some). The results of the experiments dealing with sex-limited
inheritance in Drosophila demand that we go one step further, for
they show that it is only a small part of this chromosome that
is involved in sex determination.
If this is admitted we can understand how sex may be regulated
in the same way, even when X and its mate Y appear to our rela-
tively gross methods of measurement to be equal. The differ-
ence in size between X and Y that gives a completely graded
series in different species has little to do, therefore, with condi-
tions relating to sex determination, except in so far as the initial
loss of the sex substance contained in Y led to a decrease in size.
1 am inclined to think that the difference in size relation between
X and Y represents largely the loss from Y of those materials
that play a role in sex-limited inheritance. Jf X 7s the sex chro-
mosome, then Y is the sex-limited chromosome in a double sense.
Its final disappearance in certain forms represents the total loss
of all characters that can become sex-limited in inheritance.
If it is legitimate to draw any inference from the analogy be-
tween sex determining factors, and factors that determine other
characters of the organism, it follows with a fair degree of plaus-
ibility that Y lost a sex factor (contained in the three X’s of the
species, 2 in the female, 1 in the male) in the same way that it
has lost other factors also. If these factors are, as I have sug-
gested, on a par, it follows that the material in question is the
female determining factor, F. Where then is the male determin-
ing factor? I have given my reason recently for dissenting from
the position taken by several writers that the male condition is
simply less X chromatin. It seems to me that if we treat the
problem of sex determination by the same methods used for Men-
delian characters in general, we can not Justify such a position
but are led inevitably to the conclusion that if the X-chromosome
contains (not is) the factor for producing a female, the factors
CHROMOSOMES AND SEX-LIMITED INHERITANCE 411
for producing the male must be located in some other chromosome.®
This interpretation I have developed briefly in a recent article’
in which the female factor supposed to be contained in the X-
chromosome is represented by F and the male factor, supposed
to be contained in some other chromosome (not in Y however
which is ranked with the three X’s except in so far as certain fac-
tors have been lost), is represented by M. The following scheme
shows how the relation of the sexes on this basis and how sex is
determined:
Gametes of female FM-FM a XM-XM
Gametes of male FM-M XM-M
FP female FMFM ; XMXM
, male FMM 2 XMM
56 Not in X because in males having only one X (no Y) the scheme will not work
out.
7The application of the conception of pure lines to sex limited inheritance and
to sexual dimorphism. The American Naturalist, Feb., 1911.
‘ PLATE 1
EXPLANATION OF FIGURES
Normal red-eyed fly.
Yellow fly with white eyes and “short proportionate’”’ (or miniature) wings.
Short winged fly (as in fig. 2) with normal color and white eyes.
4 Brown fly with red eyes and long wings. This fly is yellowish, but differs
from fig. 2 in the absence of a black factor (absent also in fig. 2) as well as a
second factor probably a yellow factor. Its formula is Br by. In the repro-
duction the bands and tip of the abdomen are too black. They should be more
like fig. 2 and the yellow should be more saffron. This fly was not used in these
experiments.
Wwe
412
CHROMOSOMES AND SEX-LIMITED INHERITANCE
T. H. MORGAN PLATE 1
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. Il, NO. 4. FE. M. Wallace and M. B, Abbott, Del.
413
ON THE STRUCTURE, PHYSIOLOGY AND USE OF
PHOTOGENIC ORGANS, WITH SPECIAL REF-
ERENCE TO THE LAMPYRIDAE
E. J. LUND
Bruce Fellow, Johns Hopkins University
NINE FIGURES
Introduction........ ; Pipe iS 2 ths SRO eR 415
Material. 22... a ne tea c ey teen tore ears , 416
Methods......... ; 417
Structure of the wiaatior ante organs 08: the Ls cpemidae oo 418
Physiology.......... 5 oe Sh SoS Se 431
A. Effects of ehemcals! ER a2 bn AEE SRNR CEM Ee Sead ores 437
B. Observations on the ioe alization oF the photogenic process in the organs. 439
C. Effects of rapid changes in air pressure, upon the organs......... .. 442
DR (Controlieey sa. nieiiaeiice : aerate: Meta Booopoes Ze5
E. Effects of temperature......... ae : Ae ro ae . 448
. Upon the photogenic process. bu Serta ese ete ... 448
b. Upon the photogenic Beeretion! of Cyoridina squamosa (?) and
Cyclopina gracilis..... es AS eno a: 449
ce: Uponiithetosmic acid reaction............--.:0.+: Asie 451
Use of the organs......... Be dea ene oers Lei ased aoe 455
SUMMIM Aiye eerste hres me see sett ae sara ts as Pas EPPA eR rive to eit oT Cee 457
INTRODUCTION
The subject matter of this investigation is presented in some
detail, partly because the forms of Lampyridae here studied
have not been studied before and partly because variety of mate-
rial has been available. I have attempted to bring together the
main results of other workers on the subject with my own in
order that some of the points in the anatomy, physiology and use
of these organs of light production may become more intelligible.
The literature on the subject of animal and plant photogeny
that has accumulated is vast. Yet the advance in our knowledge
of the processes, structure and use of the photogenic tissues lies
mostly within the last forty years. Often the observations by
different authors vary and are contradictory upon points that
415
416 E. J. LUND
would seem simple, and as a resuit a great number of notes and
references are found in running over the literature that are little
more than confusing and have added nothing to our knowledge
of the subject. A list of the most important papers bearing
directly upon the subject matter and others to which reference
will be made has been appended at the end of this article.
I am. indebted to Professor E. A. Andrews for his kindly inter-
est in the work and for reading the manuscript, to Professor H. F.
Nachtrieb of the University of Minnesota, and to Professor
C. M. Child of the University of Chicago for generously placing
laboratory facilities at my disposal during the summer seasons
of 1909 and 1911.
I shall use the term photogeny which has come into more gen-
eral use of late, to designate the processes concerned in the pro-
duction of light by living organisms, without rejecting nor accept-
ing the term phosphorescence as a proper one; first, because the
latter has come to stand for apparently different phenomena
occurring under varying conditions, and second, in view of the
fact that our former conception of the processes of oxidation has
undergone change in recent times and is continuing to do so.
Le Bon has employed the general term phosphorescence in making
a perhaps serviceable division of the phenomena: (1) Phosphor-
escence generated by light; (2) Phosphorescence independent of
light and determined by different physical excitants, such as
heat, friction, electricity and the X-rays; (3) Phosphorescence
by chemical reaction; (4) Invisible phosphorescence. Accord-
ing to this division our problem can with certainty be placed
under the third head, for there is abundant proof that the pro-
duction of light in animal and plant tissues is a result of some
kind of change in the chemical constitution of the substances
concerned in the process.
MATERIAL
The material was obtained during the summers of 1908-11.
The early part of the work was done upon material collected in
the vicinity of St. Paul, Minnesota. Other material was col-
lected at various points in Jamaica throughout the season, 1910,
PHOTOGENIC ORGANS 417
during the summer session of the Johns Hopkins University
Marine Zoological Laboratory located at Montego Bay, while
material used in 1911 was obtained in the vicinity of Chicago.
The time for the occurrence of the different species of Lampy-
ridae that were studied is about the same wherever they were
observed, lasting from about the last part of June to about the
first part of August in Minnesota. The period of their occur-
rence in Jamaica is considerably longer though varying with the
altitude and various climatic conditions. The following identi-
fied species of luminous Lampyridae have been studied especially
as regards the anatomy of the photogenic organs.
Minnesota specics
Pyropyga indicta Lec.
Lecontia (Pyractomena) lucifera Melsh.
Photuris pennsylvanicus DeGeer.
Photinus ardens Lec.—two varieties.
Jamaica species
Photinus maritimus.
Photinus commissus E, Oliv.
Photinus pallens Fabr.
Photinus pantoni E, Oliv.
Photinus ebriosus E, Oliv.
Photinus suavis H, Oliv.
Photuris jamaicencis E, Oliv.
Several other forms of Photinus sp. (?).
Studies on the physiology have been mostly upon Photuris pennsylvanicus,
Photinus ardens, P. maritimus, P. ebriosus. P. pallens.
METHODS
The varying methods of preparation used by different authors,
no doubt in part account for some of the minor differences in
their descriptions of the structure of the organs. Most of the
papers deal with the European species Luciola italica, Lampy-
ris splendidula and L. noctiluea. I have not had the opportunity
to get material of these species. However, the descriptions show
that no essential histological differences exist between the Kuro-
pean, American and Jamaican forms so far studied, though there
may be lesser differences, consisting mainly in the size and posi-
tion of the organs on the abdominal segments and the grosser
relations of the tracheal system to the layers of the photogenic
418 E. J. LUND
organ. In killing and fixing, the abdomen was cut off before
immersing in the fixing fluid. All of the more common killing and
fixing agents were tried. Among these formalin 4 to 10 per cent,
alcohol 80 per cent, Flemming’s fluid, osmie acid 0.5 per cent
aqueous solution, and water at the boiling point, were the most
useful, depending upon what part or structure the preparation was
intended to bring out. Other methods employed will be described
later on. All sections were stained on the slide or cover glass. A
large number of stains were tried and as a result the following were
found most useful: Thionin aqueous solution, Borax carmine
and Lyons blue, and Ranviers picro carmine as general stains.
The best of all was found to be a solution made up as follows:
anilin blue 0.5 gram, orange G. 2.0 grams, oxalic acid 2.5 grams,
water 100 ec. Instead of anilin blue Lyon’s blue may be used,
though the preparations obtained with the latter were much
inferior. Photogenic tissue fixed in boiling water for one minute
afforded beautiful preparations. The photogenic cells of Odon-
tosyllis pachydonta could be differentiated from among the other
epidermal cells. Granules of the photogenic cells; fat globules
in the fat cells, and chitin stain orange-yellow (and yellow)
respectively; cytoplasm light blue, cell membranes where present
dark blue, granules of the dorsal layer (urate granules of some
authors) stain blue. To show relations of the tracheal capil-
laries and tracheal end cells a 0.5 per cent aqueous solution of
osmic acid was used, sections cut 1 to 15 microns thick and
mounted to advantage without staining. By means of dissec-
tion of the active organs and examination under the binocular
or compound microscope in a dark room facts fundamental to
an understanding of the photogenic process could be made out.
STRUCTURE
The following description of the structure of the photogenic
organs applies to all the species of Lampyridae examined except
when otherwise stated. The location of the organs is limited
to the sternal plate of the fifth and sixth or part of either the fifth
or sixth abdominal segment. Jn all the forms studied the chitin
PHOTOGENIC ORGANS 419
of the sternae opposite the photogenic organs is transparent and
covered with hairs as are other parts of the abdominal ring. — Lin-
ing the inside of the chitinous integument is the thin hypodermis,
represented by a single layer of flattened cells with flattened
nuclei (figs. 7and 8 H). No evidence of proliferation of the hypo-
dermal cells and their subsequent differentiation into cells of
the photogenic layer has been obtained from any of the prep-
arations of the adult forms studied. DuBois has studied the
development of the photogenic organs in Lampyris noctiluca
and Pyrophorus noctilucus and states that the tissues are derived
from proliferating cells of the hypodermis. If this observation
be correct. it would preclude the photogenic cells from having a
common origin with the cells of the fat-body as some authors
have supposed. The fact that the granules of the cells of the
photogenic layer stain in several respects like those of the globules
in the cells of the fat body, affords by itself no evidence of an
ontogenetic relationship of the fat-body and. the photogenic
cells. In all the species the organ consists of the two usual layers,
the dorsal or ‘‘urate cell layer’? of some authors and the lower or
photogenic layer. The photogenic layer is completely enclosed
by the dorsal layer, except where the former is applied to the
hypodermis over the whole or part of the sternite depending upon
the size of the organ. I have found no trace of any membrane
covering the inner side of the organ as Wielowiejski (’82) states
that he found in Lampyris splendidula. Townshend (’04) finds
no membrane present in the American form, Photinus margin-
ellus, and no such structure has been found in the other European
species.
In consideration of the theories advanced by Heineman (’86)
and DuBois (95) to account for the control of the organs by
means of the segmental muscles in producing increased air pres-
sure or causing a flow of blood through the organ we should per-
haps expect to find a greater development of the muscles related
to them. This however is not the case. The muscular devel-
opment in the segments bearing the photogenic organs is not more
extended than in any of the segments anterior to them. The
muscular apparatus of each of the fifth and sixth segments con-
420 E. J. LUND
sists of three main sets of muscles, one dorsi-ventral, and two
antero-posterior sets one dorsal and one ventral (fig. 1, a, b, c).
The dorsi-ventral set is composed of two bundles one on each side
in each segment. These are inserted dorsally in the tergum and
penetrating the tissue of the photogenic organ are inserted on
the ventral side in the chitin of the sternite, appearing onthe
exterior as a visible spot or indentation which is not luminous.
These muscle fascicles are not in contact with the photogenic layer
Fig. 1 Camera lucida outline of abdomen to show relations of muscles and
photogenic organ. A, longitudinal section; B, cross section; a, longitudinal tergal
muscles; b, longitudinal sternal muscles; c, vertical muscles or ‘vertical expira-
tors;’ d, dorsal layer; e, photogenic layer; f, vertical tracheae.
but separated from it by more or less of a continuation of the
dorsal layer. The antero-posterior muscles are shown in fig. 1.
Their insertions are on the tergal and sternal membranes. They
serve to shorten the abdomen by telescoping the abdominal rings.
Their position and arrangement has no definite relation to any
single trachea or group of tracheal trunks. The tissue of the
photogenic organ does not adhere tightly to the chitinous integ-
PHOTOGENIC ORGANS 421
ument, e. g., P. pennsylvanica and P. pallens. It is of a pasty
consistency yet firm enough so that it may readily be separated
from the integument intact by proper manipulation with a small,
soft brush. In this way it may be studied with readiness under
a low or high power, treated with reagents, ete. A more direct
effect of the reagent can be observed in this way, than by simply
immersing the whole animal or abdomen. The two layers of
the organs vary in thickness within certain limits in different
species, but no marked seasonal variation of either layer has
been found. However, relative changes in the contents of the
cells of the dorsal and photogenic layers are plainly evident (p.4°2).
Wielowiejski (’89) came to the conclusion that cells of the layers
in L. splendidula were distinct. Bongardt (03) found no photo-
genic cells in the process of transformation into cells of the dor-
sal layer in the same form. Du Bois (’95) states that the cells
of the photogenic layer in L. noctiluca are transformed directly
into the erystalline deposit which he states makes up the dorsal
layer. He figures no cellular structure in the dorsal layer. Bon-
gardt (’03) also found cells in L. noctilueca which combined some
of the characters of both the photogenic and the dorsal layer
cells, and says that they may in so far be considered ‘Ubergang-
zellen.’ The figures of both these authors are too inadequate
to enable me to make out what all the facts are. In all the
species the active adults of which I have examined, the dorsal
and photogenic cells (figs. 7 and 8) are different as regards nuclei,
cytoplasm and cytoplasmic granules, and no intermediate or
transformation stages of the cells as such of one layer into the
other has been found. This was most clearly shown in sections
of Photinus ardens, P. ebriosus, P. pallens, Photuris jamaicencis
and P. pennsylvanieca fixed in osmic acid and stained in iron-
haematoxylin followed by van Gieson’s-picric-acid-fuchsin, also
with phospho-molybdie acid haematoxylin (fig. 7), thionin-eosin,
and best of all by the anilin blue-orange G. stain. In these
preparations the groups of photogenic cells show a definite pe-
ripheral boundary (fig. 7, Z.), the cytoplasm of which is nearly
uniform. Inside of this boundary are the granular contents.
In some species tracheal-end-cells of the tracheal system are
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, No. 4
422 E. J. LUND
found between the two layers and may readily be mistaken for
cells which seem to be transforming from cells of one, into cells
of the other layer, if they are not brought out by the use of osmic
acid. Cells which seem to be in the process of transformation
may also appear in sections which have been improperly and
weakly stained in, e. g., picro-carmine. In some such prepara-
tions it is almost impossible to distinguish any sharp line of separ-
ation but other sections of the same specimen when stained with,
e.g., the anilin-orange G. stain show a sharp line of demarcation
between the layers. The dorsal layer consists of closely packed
polygonal cells sharply outlined by cell walls and with a nucleus
centrally located. The cytoplasm may contain different amounts
of granule deposit, depending upon the physiological state of
the photogenic organ. The cells however nearly always retain
their original outline and size, showing that the visible difference
in the state of the cells is only that of the amount of the product
of metabolism. The filling up of the cells of the dorsal layer may
proceed to such an extent that apparently all the cytoplasm
becomes displaced, the original cell boundary and position of
the nucleus barely remaining visible. How this comes about
we shall see very shortly. In all the species the organs are pene-
trated by tracheal trunks which arise from more or less irregular
branching tracheal trunks in the body cavity or from tracheae
which lie upon the dorsal side of the organ. These tracheae which
pass into the organ penetrate the dorsal and photogenic layers
in different ways. Thus in males and females of P. pennsyl-
vanica P. pallens, P. jamaicencis and males of P. suavis and P.
commissus the vertical tracheation is most pronounced. Taking
all the species as a whole we find a great variation in the direc-
tion of penetration of the tracheae through the organs. In the
forms which have a thick photogenic organ the tracheae as
a rule are vertical while in species which have a thinner organ
or the photogenie layer only one cell deep very few tracheae
penetrate the photogenic layer. In the latter forms the tracheae
branch and spread out upon the ventral side and between the
dorsal and photogenic layers of the organ (figs. 3 and 8). The
illustrations are from such organs as have their tracheae verti-
PHOTOGENIC ORGANS 423
cally placed except fig. 8. Sections of organs with vertical or
horizontal tracheation show the general relation of parts most
distinctly.
The photogenic layer consists of three different kinds of cells
(figs. 4, 7 and 8) (a) The cells of the tracheal epithelium, (b) The
tracheal end cells,! and (c) Densely granular photogenic cells,
which make up the greater part of the photogenic layer though
this may vary in different species and different specimens of the
same species. When a horizontal section (fig. 4) is made through
an organ with vertical tracheation thereby obtaining cross sec-
tions of the vertical tracheae, we find the lumens of the vertical
tracheae appearing as round openings (fig. 4, Z). Immediately
outside of the vertical tracheal tube, crescent shaped nuclei
may be found applied close to it. (Fig. 4 NV.) These designate
the position of the cells of the tracheal epithelium and are the
only cells to be considered as making up the tracheal epithel-
ium. The cytoplasm of the’ tracheal epithelial cells is clear
and small in amount. Outside of this structure are found
somewhat irregularly placed nuclei of cells which are arranged
in a concentric ring or mass around the vertical trachea and
its epithelium. These are the tracheal end cells. They form
a vertical hollow cylinder (figs. 3, 4, # and 7, #) around the
vertical trachea and the length of this cylinder of cells shown
in section in figs. 3 and 7 is equal to the length of the part
of the vertical trachea which passes through the photogenic
layer. If the trachea assumes an oblique direction the extent
of the cylinder or tube of tracheal end cells is longer and
follows the trachea throughout the whole distance which it
passes in the photogenic layer. The distinction between the
cells of the tracheal epithelium and the tracheal end cells has
not been made heretofore. The tracheal end cells have often
been considered as constituting the tracheal epithelium. Town-
shend (’04) states that they are distinct in P. marginellus.
‘The term transition cell has been used by Townshend (’04) and Holmgren
(95). The term tracheal-end-cell however indicates their position, they being
terminal to the part of the tracheae which have taenidia (Fig. 2), andsince no nuclei
indicating the presence of a distinct tracheal epithelium is present I have preferred
te use the older term.
424 E. J. LUND
Emery, and Wielowiejski consider the tracheal cells to be derived
from the tracheal epithelium but do not state in what way they
are to be considered as being derived. To me it seems most
reasonable to consider the tracheal end cell to be a greatly en-
larged terminal tracheal epithelial cell which has come to surround
the tracheal furcation by virtue of its original position on the
tracheal branch. The number of cells making up the tracheal
epithelium is very small compared to the number of tracheal end
cells of such a dorsi-ventral branch. In many instances it is
difficult to be sure whether all the cells, the nuclei of which
appear as in (fig. 4, #) are actually tracheal end cells, i.e., whether
every one of these cells enclose within their cytoplasm the origin
of the tracheoles from a tracheal branch (figs. 2 and 8), or
whether some of these nuclei represent cells which do not surround
any such fureation but are simply placed in between the tracheal
end cells. In some species it is more difficult to tell than in
others, for the reason that the number of cells making up the
cylinder around the trachea is very much greater, e.g., P. pennsyl-
vanica and P. ebriosus, than where the cylinder is made up of a
single layer of cells. In order to make certain that all the cells
of the cylinder are tracheal end cells the number of points at
which reduction of osmic acid had taken place (90 to 100 more or
less) in preparations which showed all the furcations to contain
reduced osmic acid were counted and compared with the number
of nuclei present. The results (P. pennsylvanica) showed a very
close agreement in number. In cylinders where a single row of
cells in section are present, e.g., P. sp. near maritimus and Pho-
tinus sp. (?) (fig. 7) it is very evident that every cell of the tra-
cheal cylinder is a true tracheal end cell. The tracheae which
pass to the ventral side end here in two or more branches and
each one of these terminate ultimately in a tracheal end cell and
tracheoles. The latter penetrate upward into the lower side
of the photogenic cells. In many forms branches from the
part of the vertical trachea opposite the line of separation of the
two layers pass out between the two layers. These terminate
on the dorsal side of the photogenic layer as in (fig. 3, X), their
tracheoles passing ventrally. Consequently we may have a
PHOTOGENIC ORGANS 425
single photogenic cell or group of cells which are in section com-
pletely surrounded by tracheal end cells. Branching of the large
tracheae rarely occurs in the dorsal layer, the formation of branches
being almost entirely confined to the part passing through the
photogenic layer. I have never found tracheal end cells with
the exception of one or two doubtful cases in the dorsal layer nor
have I found tracheoles penetrating the dorsal layer in any of
the species. This points directly to the fact that the supply of
air or oxygen to the dorsal layer has no peculiar significance for
the functioning of its cells as is the case with the photogenic
layer.
The walls of the tracheae and branches are strengthened by
taenidia as in other forms. Erect hairs situated on the inside
of the larger tracheal trunks are often seen. I have not found
them present in the tracheae of the photogenic layer. Shortly
before or at the point where a tracheal branch enters the tra-
cheal end cell its diameter is diminished to about 2. (fig. 2).
In some species this diminution does not take place until the
point of origin of the tracheoles is reached. The taenidia dis-
appear at this point leaving a small tube which passes into the
cytoplasm of the tracheal end cell. This smooth tube soon divides
into two or more branches, the tracheoles, tracheal capillaries
of some authors. The point of fureation is always located near
the side of the tracheal end cell which is adjacent to the photo-
genic cell, some times only a very thin layer of cytoplasm remain-
ing between it and the photogenic cell. The nucleus of the tra-
cheal end cell is more proximal than the fureation (figs 2, 7 and
8). The diameter of the tracheole appears to be very nearly the
same in all the species, about 1.1. Since the tracheoles can
only be made to appear by means of the reduction of osmic acid
and since the amount of reduction varies (fig. 2) and depends
upon several conditions the tracheoles appear to have a greater
diameter in some preparations than in others. The diameter
varies-as nearly as could be determined from 1.1 to 2u. They
are tubular. This could be seen in very thin sections ly thick;
when light passed through, the cross section of the tracheole
appeared as a thick black ring, with a small opening. Their
426 E. J. LUND
diameter is not constant throughout but diminishes slightly as
they pass into the photogenic cell so that we have a slightly taper-
ing tube.
The tracheoles are easily shown to consist of a firm substance
different from the cytoplasm as others have found in the Euro-
Fig. 2 a, b, c, d, e, free hand drawings of tracheal end cell apparatus selected
from different preparations showing successive degrees of osmic acid reduction
upon the tracheoles, in the cytoplasm and distal end of trachea; n, nucleus; t,
taenidia.
pean forms and Townshend ’04 in the American form P. mar-
ginellus. Tissuetreated with a n/s. of NaOH show the tracheoles
to remain intact while the cytoplasm of the cells is dissolved away.
The structure which remains after treatment with the hydroxide
PHOTOGENIC ORGANS 427
cannot so far as staining reactions are concerned be distinguished
from the substance of the trachea and may in so far be considered
of a chitinous nature. That the substance of the tracheole is
different from that of the trachea by reason of the fact that osmic
acid is reduced upon it is of no significance, for reduction is not
always limited to the tracheole.
Weilowiejski and Bongardt (’03) found the number of trache-
oles in L. splendidula arising from each tracheal end cell to vary,
there being sometimes as many as six or seven. Emery found
uniformly two present in Luciola italica. Townshend found them
to vary in number in Photinus marginellus though usually two
were present. I have found two in the following forms, male
Photuris pennsylvanicus, Photinus sp. near maritimus, P. ebrio-
sus, P. ardens var (?). In females of Photuris pennsylvanicus I
found three present. In some cases four seemed to be present
in females (?) of P. pennsylvanicus, but this could not be defi-
nitely determined because of the character of the osmic acid reduc-
tion in these preparations. In males of Photuris pennsylvanicus,
Photinus sp. near maritimus, P. ebriosus, and P. ardens var (?)
the number is constant.
Emery considered the tracheoles to pass between the photogenic
cells in Luciola italica. Wielowiejski was of the same opinion
for Lampyris splendidula, though later (89) he found that some-
times the tracheoles did penetrate into the cells. Bongardt (’03)
states that he has not found the tracheoles to enter the photogenic
cells in any of the species which he studied. Townshend (’(04)
states that it is ‘probable that the tracheoles in their course out-
side the cylinders’ are intercellular, and Watase (95) also states
that they are intercellular. Before we attempt to answer the
question as to whether the tracheoles pass into or between the
photogenic cells in the species studied we may consider the struc-
ture of the photogenic cell.
These are shown in (figs. 3, 4, 7 and 8, P). In nearly all the
species a cell wall is absent and in some species cell boundaries
are missing. Thus the photogenic cells may be considered to
be masses of protoplasm in which nuclei are placed, and lying
between the cylinders the latter formed by the tracheal end cells,
428 E. J. LUND
The only means of distinguishing the structural individuality
of the photogenic cell where it does appear is the fact that densely
granular deposits in the cytoplasm tend to group themselves
about the region of a nucleus (fig. 4), thus leaving narrow radiat-
ing areas or strands of cytoplasm which may in some species
and apparently under some physiological conditions of the photo-
genic cells appear as cell boundaries or more rarely as cell walls,
consequently they may be spoken of as cells. The nuclei are
placed in the central parts of the cells and vary in size and shape.
They are larger than the tracheal end cell nuclei. In Photinus
commissus the nuclei are narrow and appear in rows as Townshend
figures for P. marginellus (fig. 3). In P. sp. near maritimus, for
example, they appear rounded and lie end to end forming a chain-
like vow (fig. 5). The cytoplasm of the photogenic cells are in
all cases filled with granules. In Photinus sp. (?) (Jamaican)
(fig. 7, Z) a peripheral zone of very finely granular cytoplasm
appears. The characteristic granules of the interior are missing
from this zone. I have found the same but narrower zone present
in sections of Photinus ebriosus and also in some specimens of
P. sp. near maritimus. Whether it appears as a result of partial
exhaustion of the contents in the metabolism of the cell or whether
it is a permanent condition I have as yet been unable to deter-
mine. The anilin orange G. stain most clearly differentiates the
parts of the photogenic cell. Coarse granules appear yellowish
orange while the peripheral zone takes a dense blue stain. That
these differences in the appearances and relations of the photo-
genic cells in the different forms is not due to preparation is shown
by the fact that they appear the same in sections with different
methods of fixation and staining.
To return to the capillaries; in spite of the fact that the above
writers with the exception of Wielowiejski (’89) agree that the
tracheoles are intercellular in the European forms and the Amer-
ican form studied by Townshend, I have become convinced after
examining a great many preparations that the tracheoles are in
no ease limited to the cell boundaries where such are present.
The fact that they do penetrate into the cytoplasm is clearly
shown by the fact that cross sections of the tracheoles appear
PHOTOGENIC ORGANS 429
close to the nuclei of the large photogenic cells in the same focal
plane. Furthermore in tissue where no cell boundaries are evi-
dent we find them penetrating in all conceivable directions.
Thus several osmic acid preparations of the photogenie organ of
P. ebriosus show the cytoplasm of the photogenic cells pene-
trated in all directions by an almost innumerable number of
tracheoles. Such preparations can not but convince one of the
importance of this greatly developed tracheolar net work for the
functioning of the photogenic layer.
Observations differ also as to whether or not the tracheoles
anastomose. Emery states that they do not anastomose in
Luciola italica, Wielowiejski found the tracheoles to anastomose
in some cases. Bongardt (’03) found no anastomoses. Town-
shend has figured the anastomoses in P. marginellus (figs. 5 and
6). In horizontal sections of Photinus sp. (?) (Jamaican) I
have found structures like that figured by Townshend in fig. 6.
The radiating structures arising from the tracheal end cells are
undoubtedly tracheoles. I was however unable to convince
myself that anastomoses did occur in this case. In the photo-
genic organs of P. ebriosus the tracheolar network is especially
dense and in sections showing strong reduction of osmic acid a
great many anastomoses could be plainly seen. I have found no
branching of the tracheoles of the photogenic organ in any of
the species examined. Anastomoses of tracheoles have been
found by Wistinghausen, and Holmgren in certain caterpillars.
Wistinghausen and others have found the tracheoles to be made
up of a chitinous tube with a ‘peritracheal membrane’ which is
continuous with the cytoplasm of the transition cells (correspond-
ing to the tracheal end cells of photogenic organ). It is readily
seen that anastomoses of the tracheoles renders the mechanism
of respiration much more efficient due to the possibility of flow
of the air through the tracheoles caused by unequal pressure on
different parts. Since it is found that the tracheoles do anasto-
mose in those instances where they have been made a special
study we should perhaps expect to find them filled with air. But
such however appears not to be the case so far as direct obser-
vation shows. I separated the organ of P. pennsylvanica from
430 EK. J. LUND
the sternite by means of a very small brush and immediately
examined them under the microscope in the dark. The ventral
ends of the vertical tracheae appeared as glistening, white, branch-
ing structures and were filled with air. No trace of tracheoles
could be seen though they occur upon or very near the ventral
surface of the photogenic layer, neither could they be seen by
means of ordinary light. Thus it would appear that they were
filled with liquid. From the fact that the tracheoles are found
to anastomose it becomes easier to explain this condition as being
a temporary rather than a permanent one for the tracheoles could
be more readily freed from the liquid by the unequal pressure
caused by the respiratory muscles than they could if the tracheoles
ended blindly. Bongardt (’03) found the tracheoles to be sur-
rounded by a continuation of the cytoplasm of the tracheal end
cell. Wistinghausen and Holmgren found a peritracheal ‘mem-
brane’ around the tracheoles continuous with the ‘transition
cells’ of certain caterpillars. I have not found any distinct
cytoplasmic membrane surrounding the tracheoles. However
when osmic acid becomes reduced the greater the reduction the
greater is the outside diameter of the tracheole. Furthermore in
teased preparations the cytoplasm appears spread out between
the tracheoles in a membrane like fashion and passes out upon
them a short distance (fig. 2) and as others have figured. The
reduction of the acid occurs around the fork in the cytoplasm of
the tracheal end cell and spreads outward until the whole of the
tracheal end cell becomes blackened (fig. 2). Now since in such
preparations there is always a definite limit to the amount of reduc-
tion around the tracheole it would appear that we have a limited
region around the chitin tubule which reacts to osmic acid the
same as the cytoplasm of the tracheal end cell, and hence is spe-
cifically different from the adjacent cytoplasm of the photo-
genic cell. Thus we may conclude that whatever the function or
property of the tracheal end cell may be the thin layer correspond-
ing to the peritracheal membrane of Wistinghausen and others
possesses the same properties as the cytoplasm of the tracheal end cell
so far as the osmic acid reaction indicates. The acid is not reduced
in the chitin of the tracheae, neither does the cytoplasm of the
PHOTOGENIC ORGANS 431
tracheal epithelium of the photogenic organ show this specificity.
So that whether we consider the substance of the tracheole to be
different from the chitin of the trachease and endowed with the
specific property of reducing osmie acid or whether we consider
the reduction due to the specific property of the cytoplasm of
the tracheal end cell the fact remains, that the nucleus and clear
cytoplasm of the tracheal end cell and also the outer parts of
the tracheole have the same property, viz., the power to reduce
osmie acid and that this property is primarily peculiar to the tra-
cheal end cell.
When the cells of the organ are treated with osmic acid they
of course are killed after a short time though the process of photo-
genesis may go on for some time after immersing the organs.
This is undoubtedly due to the characteristic slow penetration
of osmie acid, for I found that the light from a solution of the
luminous secretion of the ostracods Cypridinia squamosa(?),
and Cyclopina gracilis was instantly extinguished when a small
quantity of osmie acid solution was added. It is evident that
the osmic acid, reaction is not necessarily due to, and it is believed
will be shown not to be dependent upon, the temporary nor con-
tinued vitality of the cytoplasm but that it is due to some spe-
cific formed substance present in the cytoplasm of the tracheal
end cell and the peritracheolar membrane.2 To this question we
shall return later.
PHYSIOLOGY
As stated above the crystalline deposit in the cytoplasm of
the dorsal layer cells may vary in amount. I have found that
where the crystals appear in abundance the photogenic cells
are shrunken and the region of the tracheal cylinders are large and
vacuolated, while in cases where the crystalline deposit is small
or not present the photogenic cell groups are plump and the cyto-
plasm very densely crowded with granules. That this is not due
to methods of fixation, staining, etc., is clearly shown by the fact
2 It seems to me that the proper term for this membrane should be peritracheolar
membrane rather than the misnomer ‘peritracheal membrane’ since it is peculiar
to the tracheole rather than the trachea.
432 E. J. LUND
that the same differences appear in material carried through
different processes of preparation. In extreme cases the filling
of the dorsal layer cells has gone on to such an extent that the
cell boundaries have almost disappeared. When the tissue is
treated with gold chloride and formic acid the granules appear
in the form of large crystals in the dorsal layer. These are either
monoaxial or biaxial crystals. The granular deposit in the dor-
sal layer under normal conditions where only alcohol and xylol
have been used in preparation of the sections appears uniformly
granular like it appears in fresh tissue. The same is true when
osmic acid has been used on such organs. The granules are
minute crystals less than 0.54 in diameter. Their crystalline
nature is shown by their high birefringency. They are totally
different from the granules of the photogenic cells which are
non-crystalline and yield totally different reactions. Bongardt
(03) and Koélliker (64) found them to be ammonium urate and
after adding NaOH ammonia was given off. Others have found
them to be ‘some salt’ of uric acid. When the xanthin testis
applied to the isolated tissue, the residue after treatment with
concentrated HNO; and ammonia vapor shows a reddish purple
color. Upon subsequent addition of NaOH however, the color
turns deep reddish-brown instead of bluish-violet. The tissue
gives the test for guanin under some conditions, with an alkaline
solution of diazobenzolsulphonic acid. It would appear from
these results that we have here a nitrogenous compound related
to if not identical with some of the derivatives from nucleic
acid. The interesting point is that we would accordingly have
nitrogen present, and perhaps phosphorus. This immediately
suggests that we should expect the granules of the photogenic cells
to be a nitrogenous compound since a direct relation exists between
the amounts of the granular contents in the layers. Furthermore
this direct relation and the actual tracing of the products of decom-
position resulting from the process of photogenesis from their
place of origin in the photogenic cell into the dorsal layer cells,
is shown strikingly and conclusively in sections of the organs of
different species (e.g., P. pennsylvanica, P. ebriosus) prepared
by a variety of methods. The photograph (fig. 5) shows a view
PHOTOGENIC ORGANS 433
of a longitudinal section of the photogenic organ of P. sp. near
maritimus. The dense white mass is the crystalline deposit in
the dorsal layer cells, in this case the cells are completely filled
with the minute erystals. Above this mass is shown small
amounts of the same product among the viscera of the body
cavity. On the ventral side shown under high power in fig. 6
are the same products definitely located on the most peripheral
parts of the cytoplasm of the photogenic cells, between the tra-
cheal end cells making up the cylinder and the photogenic cell.
It is important to note the exact limits in the location of this
crystalline deposit none being found within the cytoplasm of
the tracheal end cell nor in any part of the tracheal cylinder.
The outlines of the nuclei are shown in (fig 5) appearing in chains,
and separate in (fig. 6). They are surrounded by a layer of the
same substance. The photographs are from specimens which
showed the greatest amount of accumulated products of photo-
genesis in the organ found in any of the material examined. I
have found different amounts of the accumulated products of
katabolism in different species and different specimens of the same
species, so that we may find organs with only a small amount of
deposit on the periphery of the photogenic cell masses with no
such deposit upon the surface of the nucleus and in many cases
if not in most cases very little if any exists in the photogenic layer.
The degree of filling of the dorsal layer cells also corresponds to
the amounts of the deposit upon and in the photogenic cells.
It is important to note that the region of the nuclei is the last
to show the presence of the waste products. This would seem to
indicate that the chemical processes that have to do with the for-
mation of this crystalline decomposition product can be traced
back to the region of the nucleus. This point becomes more
interesting when we compare the results obtained from the inves-
tigations by R. 8. Lillie (02) and others on the properties and
functions of the cell nucleus in metabolism, with special reference
to its oxidative properties. From what follows it will be seen
that the foregoing may or may not have to do with the immedi-
ate processes of photogenesis, i.e., determine the localization of
the origin of the light emissions from the parts of the photogenic
434 Hy) Jie LUND.
cell. It is to be considered as a condition where the waste prod-
uct has not been removed from the photogenic cells, but has
begun to accumulate. In this condition the organ is still fune-
tional and the intensity so far as can be seen with the naked eye
is apparently the same. From the facts concerning the rela-
tion, In amount, of granular deposit in the dorsal and ventral
layers it is evident that we are to consider the granules of the
photogenic cells as at least one if not the main source from which
the crystalline deposit is derived. Furthermore DuBois states
that he observed in the luminous secretion of the centipede, Orya
barbarica, a direct transformation of the granules of the photo-
genic secretion into crystals. From this it would appear that
Weitlaner (’09) is in error when he concludes (p. 103) that ‘‘3.
Die harnsauren Ammoniakschéllechen Ko6llikers [crystalline de-
posit] haben den Hauptteil am Leuchten, sie sind die Elemente
des Leuchtens und man spricht richtiger von Leuchstoff als von
Leucht-organen,’’ and in view of the fact that observations by
other workers upon the forms which he studied (Lampyris splen-
didula and L. noctiluca) and also as far as observation on other
forms go, the dorsal layer is non-photogenic. I have found no
traces of ight from any parts of the body cavity, and the fact
that small amounts of the crystalline substance do occur in the
body cavity is of course no evidence for the localization of a proc-
ess of photogenesis in the body cavity or fat body since the sub-
stance is soluble and in part undoubtedly dissolves in the fluids
of the body cavity. We are not even justified in concluding that
the origin or localization of the light in the photogenic tissue is
exactly the parts of the photogenic cells which as shown in (fig 6)
contain the crystalline deposit, for observation shows that light
isnot limited in its origin to the regions where the crystalline
deposit appears in the photogenic cell. All we can say is that the
places in the photogenic cell where the crystalline deposit appears
in preparations, are the points at which crystallization takes place
or where the cytoplasm becomes most saturated with the deeompo-
sition product. It is of vital importance to one’s understand-
ing of the processes of cell metabolism that we do not simply
PHOTOGENIC ORGANS 435
assume simplicity and let this assumption cause us to overlook
evident complexity in the processes which go on in the cell.
The granules of the photogenic cell are uniform in size approxi-
mately 0.4 in diameter. In fresh tissue they do not appear to
be spherical. The fat globules in the cells of the fat body are
from 5 to 15 times larger and perfectly spherical. Thus we see
at once a difference between them in some of their physical
properties. In some of the literature on the subject when an
explanation of the process of photogenesis is attempted it is usu-
ally stated that the process is ‘‘probably one of oxidation of a
substance of fatty nature.’ This led me to try various micro-
chemical tests for fats and related substances. Osmic acid has
usually been regarded as a specific ‘stain’ for fat. Lately Sudan
III and Scharlach R. have been used and considered to be more
accurate tests. Osmic acid blackens the globules in the fat body
under all conditions, so that it becomes of importance to deter-
mine if possible the conditions under which this substance is
and is not reduced in the tracheal end cell where no such fat deposit
is present but instead a clear cytoplasm. To this we shall return
later.
In order to see what the effect would be to feed the insects
with Sudan III as has been done by Riddle (710) in studying the
deposition of the yolk layers in the hens egg I fed specimens of
the various species and also Pyrophorus plagiophthalmus on sugar
cane impregnated with Sudan III for five or six days. The ani-
mals remained perfectly healthy and active. Such animals became
pinkish in color showing that the Sudan III was distributed
through the body tissues. Sections showed that the stain had
been strongly deposited in the fat globules of the fat cells, the for-
mer being stained a deep orange red. A faint pinkish coloration
took place in the other tissues, which was very probably due to
the presence of the stain in solution in the body fluids.. The
photogenic layer showed a slightly stronger stain than the other
tissues of the body with the exception of the fat globules in the
fat cells. But with such a slight coloration it is impossible to
tell whether this slightly stronger staining in the photogenic cells
than in the body may not have been due to the smaller size and
436 E. J. LUND
greater number of granules of the photogenic cells, and hence it
is difficult to say whether slightly weaker color was due to smaller
degree of specific affinity for the stain than in the case of the fat
globules or as simply due to the physical conditions. This at
once indicates a difference inthe chemical nature of the fat deposit
in the fat cells and the granules of the photogenic cells. Since
as has been pointed out above and has also been found to be
true by others that the waste product from the photogenic cells is
a nitrogenous compound, we should expect nitrogen present in the
substances used up by the photogenic cell in the processes of
photogenesis and since many of the ordinary stains show similar
reactions of the globules of the fat cells and the granules of the
photogenic cells, they must be considered as having some proper-
ties in common. This led me to compare as far as possible the
reactions of the nitrogenized fats with the reactions of the gran-
ules of the photogenic cells.
Loisel (’03) has studied and compared the staining reactions of
neutral fats with those of lecithin by a method which he recom-
mends (q.v.) and has found marked differences in their affinities
for many stains. By mordanting with iron-alum, staining for
a short time and then dehydrating with acetone, the following
results are obtained.
STAIN PHOTOGENIC GRANULES FAT GLOBULES
Aqueous sol. orange G.. yellowish-orange not stained
Methyl green... bluish-green pale green
Acid fuchsin.... deep purple very pale purple or not stained
Gentian violet deep violet very faintly violet or not stained
Erythrosin deep crimson not stained
Also
Phosphomolybdic-hae- dark blue-black | not stained, clear
matoxylin (a myelin |
stain)
Sudan III (fed)...... slight or none.’ | deep orange red
3 This agrees with results of Daddi on lecethin and fat. Archiv, ital. de Bio-
logie, ’96. t. 26, p. 145.
PHOTOGENIC ORGANS 437
These staining reactions agree with those for lecithin and fat
respectively found by Loisel and Daddi. Their differences are
so marked in all cases that if the staining reaction is to be regarded
as an index to the chemical nature of a tissue substance at all it
seems that it must be regarded so in this case. Of course we
are not justified in stating that the photogenic cell granules are
a lecithin until further conclusive tests can be applied. Yet it
is suggestive that a substance, the derivatives from which contain
nitrogen so closely resemble the well known lecithins in all the
reactions that it has so far been possible to obtain.
Effects of certain chemicals
The dorsal layer of the photogenic organ shows a strongly acid
reaction while the photogenic layer has a slightly weaker acid
reaction. This was tried several times by removing the two lay-
ers from the transparent chitin and then turning the respective
layers down upon the litmus paper and comparing the resulting
colors. Heineman also found the tissues of the organ of Py¥ro-
phorus to have an acid reaction. This of course would not neces-
sitate that the photogenic process take place in an acid medium.
for it is well known that different parts of the same cell and the
same parts of a cell at different times may give different reactions
as regards acidity and alkalinity. Watase (95, p. 115) reason-
ing from analogy to certain other photogenic processes which do
occur in alkaline media, states that here the process also takes
place in an alkaline medium. However, so far, all the evidence
we have, rests only upon analogy. Kastle and MeDermott (’10)
have repeated and added a large number of experiments on the
effects of chemicals on the light production of the organs of P.
pyralis. The results of these authors and many others show con-
clusively that oxygen is a necessary element for the process of
photogenesis. This statement may not be the same as saying that
the process is one of simple oxidation. J have found a strong solu-
tion of hydrogen peroxide to have a most striking effect of increas-
ing the light intensity. The following are notes from one experi-
ment :
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. LI, No. 4
438 E. J. LUND
(A) 1. The organ of a @ P. pennsylvanica was isolated
intact and inverted on a slide. One to two drops of strong H.O,
was added.
Result: A strong constant light appeared and oxygen was
evolved.
2. ‘The viscera of the body cavity were then treated with H.Q.
Result: No light but evolution of oxygen.
(B) 2. A second organ was isolated in the same way and
placed on the slide dorsal layer up then H.O, added.
resull: Bright, strong, constant light appeared. The dorsal
layer was opaque to light. When one of these brightly glowing
organs were placed in the tube and bulb apparatus (see p. 443)
and pressure increased a slight increase in intensity of the light
could be noted although the organs were apparently at their
upper limit of normal intensity. When left moistened with H.O,
in the air they remained brightly luminous for two hours. The
results of this experiment and from the results found by Kastle
and McDermott (’10), Bongardt (03) and others that in pure
oxygen the light is more intense than under conditions where
oxygen is diluted as in air, show at once that a direct relation
exists between the intensity of the light and the oxygen content
or pressure. Kastle and McDermott and other previous writers
have shown that water is a necessary condition for the existence
of the ight and that by drying the tissue of the organs after lorg
periods (thirteen months in one case, Kastle and McDermott)
light may be emitted from the substance upon admittance of
water and oxygen. These and other writers have further found
that by repeated drying and moistening with the admission of
oxygen the process of photogenesis ceases after a time indicat-
ing that the photogenic process is a result of a direct utilization
of material. This is of course in perfect agreement with the
results reported above on the origin of the nitrogenous waste
product in the dorsal layer, and the diminution in the cell contents
of the photogenic cells. From a study of the dried photogenic
material we at once see as others have long ago pointed out that
the immediate process of light production is not dependent
upon the protoplasm of the cell but upon interactions between
formed substances.
PHOTOGENIC ORGANS 439
Observations on the localization of the photogenic process in the
organs
While collecting at night I found that certain Jamaican species
normally emit the light in a regular way, e.g., Photinus sp. (?)
(a small yellow form), emits what appears to be a long sharp
flash lasting from 2 to 23 seconds when seen at a distance of a few
rods. This flash when examined more closely is seen to consist
of a number of smaller flashes caused by regular, rapid and numer-
ous changes in intensity. In some other species, e.g., P. sp. near
maritimus the flash is similarly made up of a series of rapid changes
in intensity but not as uniform and definite as in the above men-
tioned form. Photinus ebriosus shows a strong uniform flash.
These species may readily be recognized by the character of their
flash. A further study would undoubtedly enable one to dis-
tinguish other forms by this means. Similar observations have
been made upon some of the American Lampyridae by McDermott
Calo)
Max Schultze (64-65) from his study of the reduction of osmic
acid in the tracheal end cell suggested that the process of light
production took place in these cells. Emery (’84) was the first
to make a study of the localization of the light in the photogenic
layer. He found that the light in Luciola italica was localized
in rings which were constant in position but sometimes were
broken up into luminous points and since he found the reduced
osmic acid in the forks to be located here and in the tracheoles
he was of the opinion that the region where the photogenic proc-
ess took place was where the osmic acid was reduced.
I have studied Photuris pennsylvanicus, Photinus ebriosus, P.
pallens, P. sp. near maritimus, and others. When the organs are
at the height of their intensity and when the light is not intense
enough it is impossible to see anything with the compound micro-
scope in the dark. But when the organs show a medium inten-
sity the surface of the organ and the removed tissue can easily
be studied. A view (fig. 9) under the high power, of the organs
4 Unfortunately this form could not be identified for me.
5 Can. Ento., vol. 42, pp. 357-363.
440 E. J. LUND
of, e g., P. sp. near maritimus shows the surface of the photogenic
layer to be made up of a number of shaded, oval or round spots
distributed uniformly in a brightly luminous field. This generally
occurs when the animal is under constant mechanical stimula-
tion such as very slight pressure or when the organ is removed
and then stimulated. To identify these shaded areas camera
lucida drawings were made of the active organ and then horizontal
sections cut and compared. These showed that the shaded areas
(fig 9, S) corresponded exactly to the cross sections of the vertical
cylinders and the bright areas between them to the photogenic
cells as in (fig. 4). Thus from direct observation we are readily
able to locate the origin of the light in the cytoplasmic region of
the photogenic cells. In certain favorably lighted regions on
the organs, the yellowish green light is most intense in the region
which would correspond approximately to the inner region of
the zone bounding the photogenic cells (figs. 7, Z and 9, P). This
zone 1s narrow in most forms and almost absent in some so that
as far as could be determined the periphery of the photogenic
cell shows under these conditions the greatest intensity. It must
be noted here however, that in many cases sections show that
the tracheal end cells lie in depressions in the cytoplasm of the
photogenic cells as is shown in an extreme case in (fig. 8) so that
we should expect to have a more or less indefinite bright zone
about the vertical cylinder if the light is located in some definite
part of this region. The nuclei of the photogenic cells appeared
in bead-like rows in some places. The intensity of the light was
no greater in the region around the nucleus where the crystalline
deposit appears than in the other regions of the cytoplasm, thus
the place where the waste products originate can not be consid-
ered to be limited to the places where they crystallize out for we
must assume that the nitrogenous waste product is at least par-
tially formed at the points of origin of the ight. Under some con-
ditions, often when the tissue of the organ is injured, the light
proceeds from certain angular small spots in the form of an intense
bright flash. This may occur a great many times at regular
or irregular intervals. The angular spots are always constant
in position and shape but the intensity of the flash may vary.
PHOTOGENIC ORGANS 441
I attempted to see if there was any regular distribution of them
but none such could be made out definitely. During the emis-
sion of the flashes from these points clouds of diffused light orig-
inating from the deeper parts of the photogenic layer often passed
over the organ. These generally proceeded from areas similar
if not identical in outline and constant in position. The organs
in this condition of activity appear very much like the view one
may obtain in the spinthariscope. When the same organ was
stimulated the total light content increased greatly the charac-
teristic shaded areas with the light regions between them reap-
peared, changing the general view of the field. Thus there
appears to be two modes of photogenic activity of the organ which
show that the mutual relation of the factors which make the
photogenic process possible is not a constant and fixed one in all
respects but that through some variation in these relations we may
have the location where the photogenic reaction takes place limited
to certain special regions under some conditions while under other
conditions the localization may be extended in area. This is a
‘spreading’ of the process and a result of increased stimulus.
Structurally it is impossible to refer these angular points
which are constant in position to anything except the region of
the forks of the tracheoles in the very distal part of the tracheal
end cell, as Emery did. The fact that those points which were
visible, were placed irregularly is very likely due to the distri-
bution of those forks which are at the ends of the branches of
the vertical tracheae on the ventral surface for here the forks are
scattered irregularly over the ventral surface of the photogenic
layer (fig. 3). The fact that the greatest intensity of the light
when such that the organ can be examined is located on the
periphery of the photogenic cell would indicate as Emery also
found that the primary liberation of light energy is from the
region where the forks surrounded by their thin layer of tracheal
end cell cytoplasm are imbedded in (fig. 8) or applied to (figs 4
and 7) the cytoplasm of the photogenic cell. When we consider
this to be the case the fact that upon increased stimulation, the
cytoplasm of the photogenic cell appears to be the origin of the
light, becomes easily explainable for through this we have innu-
442 E. J. LUND
merable tracheoles passing, and if the photogenic process took
place around them the cytoplasm would be uniformly illumined
as is found to be the case. On the other hand if we assume the
photogenic process to actually take place uniformly throughout
the cytoplasm of the photogenic cell we would have to find an
explanation for the bright angular points in the less stimulated
state of the organ. The tracheal end cells are not luminous and
the fact that the cylinders are not totally dark is due to diffusion
of light into them from the photogenic cell regions. For if we
are to attribute any degree of luminosity to them (i.e., the small
amount which they show) we should expect them to be brightly
luminous, which is not the case. Now since these cells are not
luminous and the peritracheal membrane around the tracheoles
is blackened in the same way as the tracheal end cell we have no
reason to believe that light production is localized in this mem-
brane and not in the cytoplasm. On the contrary we should
expect that the process of light production really takes place
in the cytoplasm of the photogenic cell and very close to the con-
tact surfaces between the tracheal end, and photogenic cells
of the organ. This leads us farther to localize the angular spots
in the cytoplasm of the photogenic cell immediately adjacent to the
forks and tracheoles, for here we evidently have the greatest amount
of reaction because of the presence of the fork and body of the
tracheal end cell at this point. This agrees with all the facts
from observation. The same explanation was arrived at from a
detailed study of the osmic acid reaction before the above study
of the active organs had been made. Finally when we come
to consider the control of the organ the above conclusion would
seem to be more in accord with the facts.
Effects of rapid changes in pressure wpon the organ
DuBois (’86) found that under diminished air pressure the
light became weaker in Pyrophorus and at six hundred atmos-
pheres the organs were intensely active. Evidently changes
in air pressure in the tracheae normally vary only between narrow
limits and if these changes have anything to do with the control
of the organ we should expect to be able to affect the action of
PHOTOGENIC ORGANS 443
the organ by rapidly changing the air pressure in the tracheae
of the body and photogenic organ. In order to approach this
condition as nearly as possible a strong rubber bulb of 50 ce.
capacity was fitted to one end of a heavy rubber tube at the other
end of which was fitted a 2 dram vial. The specimen was then
placed in this vial and a plug of slightly moistened cotton wool
was inserted so as to keep the animal in the tube. The following
are the notes taken in one experiment from among many on differ-
ent forms.
Photurus pennsylvanica 3. (a). The actively flashing ani-
mal placed in the vial and pressure applied with the hand.
Result: Flashes normally bright and strong when under atmos-
pherie pressure. Increase of pressure increased intensity of flash
temporarily.
(b). I then cut off the abdomen from the same individual.
Light became totally extinguished. On placing it in the vial,
no light was produced at atmospheric pressure. I pressed the
bulb and as the pressure was increased light appeared more
intense in proportion to increase in pressure. An intense flash
could be produced in this way though not quite as intense as
that of the normal animal. Flashes were produced at will, at
intervals, for twenty-five minutes. Only a very slight diminu-
tion in the intensity of the light at the end of twenty-five minutes
was noted.
(ec). I then isolated the organ from the same individual com-
pletely from all other tissue and placed it in the vial. A faint
constant light was evident. I increased pressure and increase
in light intensity in the whole organ took place. This could be
repeated again and again, an intense flash being produced at will.
(d). I then cut the same isolated organs into halves longi-
tudinally. Effects of increase in pressure on halves was the
same as in (¢).
(e). LI removed photogenic layer from one-half of the organ in
(d) and placed the tissue in the tube. A weak constant light was
given off due very likely to stimulation of tissue. Increase of
pressure gave same results as in (c) and (d). After a short time
the tissue lost its power of response and light soon became extin-
guished.
444 E. J. LUND
It will be noted that in this experiment the increase of pressure
on the exterior was presumably the same as in the interior of
the tracheal tubes so that the only apparent effect was to produce
increased partial pressures of the oxygen and other constituents
of the air. Continued constant pressure produced corresponding
constant intensity. The conditions are not the same as when
contraction of the abdominal muscles takes place while the spir-
acles are closed, yet increase in concentration of oxygen is brought
about in the tracheal tubes assuming that a free flow of air into
the tubes takes place. Of this there can be no doubt, for the
tracheae are kept distended by the taenidia. Thus it seems that
the only explanation of these experimental facts, providing increase
of pressure does not mechanically or otherwise stimulate the
photogenic tissue, is that this increase in intensity is directly
due to the increase in oxygen content in the photogenic tissues
for as has been shown by Kastle and McDermott, and others,
the intensity of the light is greater where the oxygen content is
higher. It will be noted that these results are obtained where
the total pressure is constant. Now when the pressures are
lowered by means of suction on the rubber tube the opposite,
when any, are produced, viz., the intensity of the light is decreased
with decrease of pressure and often may be temporarily extin-
guished. The same results from experiments with low pressures
have been obtained by DuBois for Pyrophprus. In order to
eliminate the possibility that the increase in pressure is a stimulus
upon the cells in a mechanical sense I need only cite the experi-
ments of Bongardt (’03, p. 31). He took fresh and dried photo-
genic tissue of L. noctiluca (the former responded to stimuli but
the dried tissue could not be made to respond under any condi-
tions) placed them in a glass tube and slightly moistened the
dried tissue which became luminous. Then the pressure in the
tube was decreased to 10 mm. of mercury. After 243 minutes
the light totally disappeared in both the moistened and fresh
tissues. When air was again admitted (equivalent to increase in
pressure or oxygen content) both the moistened and fresh tissue
became intensely luminous. This leads us to the question of the
PHOTOGENIC ORGANS 445
Control of the organ
There have been three main theories concerning the control
of the photogenic process. First, the theory of DuBois, which
holds that the respiratory processes are only related to the con-
trol of the organ in so far as they serve to supply the necessary
amount of oxygen in the same capacity as to the other tissues
of the body. In other words that the respiratory muscular mech-
anism does not determine when a flash shall or shall not take place,
DuBois (’96) considered the control to be nervous in so far as
the nervous system controlled the action of the muscles which
are related to the photogenic organs; and that by contraction
and relaxation of these muscles blood is made to flow ‘through
passages’® in the organ. By this means of the control of the blood
flow he explains “why sensorial or psychic stimuli may affect
the production of light’? (p. 420). He notes however that the
‘photogenic cells’ are ‘directly excitable.’
The second theory is that of Heineman (86) and Watase (’95)
who hold that the control is by the respiratory muscular mechan-
ism causing the inflow of air to the photogenic tissue. Heineman’s
theory rests upon the fact that he found muscles passing over
the surface of the abdominal organ of Pyrophorus under which
passed large tracheal tubes penetrating into the photogenic tis-
sue. When he blew through a tube which was inserted into the
large ‘prothoracic spiracle’ (?) the light from the organs increased
in intensity. This he considers an ‘experimentum crucis’ sup-
porting his theory of control. The third one is that the organs
are under direct nerve control. Max Schultze (’64) considered
it probable that nerves connect with the tracheal end cells.
Bongardt (03) figures and states that he found nerve fibers passing
along the vertical tracheae and finally ending in the tracheal end
cells of L. splendidula. If such nerves do exist in the organ,
it becomes of vital importance in view of the results from the
experiments with pressure, to see if their presence can be demon-
’ These ‘passages’ which DuBois erroneously figured and considered blood
spaces in L. noctiluea are of course nothing but the regions of the vertical cylinders
and tracheae which sometimes do not stain readily and thus appear as if spaces
are present when actually no such spaces exist, as Bongardt and others have shown.
446 Ba Jd. LUND
strated by a physiological method and if so what part they play
in the process of photogeny in the Lampyridae. If we do find
such a control we will have found nothing new so far as photo-
genic organs in many other forms are concerned, for instance in
the photogenic Pennatulidae, Panceri noted the passage of the
wave of light over the colony of polyps, serving as an index of
the fact that the passage of a nervous impulse precedes the pro-
duction of light. Asto whether the result of this impulse partakes
of the nature of a contraction in the photogenic cell we shall speak
presently. Here we are particularly concerned with the question
as to whether the nerve impulse ends directly in the photogenic
tissue, and if so where it ends, or whether it ends in the parts
of a mechanism which is external to the photogenic tissue, such
as the muscles of the body.
First of all it must be noted that the respiratory movements
of the abdomen do not at all follow or correspond to the flashes
emitted from the organ. Hence the simple respiratory move-
ments do not account for the periodic emission of the light. In
trying to carry out the idea of the compressibility of the volume
of air and its translation in the tracheae of the photogenic organ
I made repeated efforts to see if any minute movements of the
tergae or other parts could be seen in the animal emitting flashes
of ight under a binocular. No trace of such movement could be
found. Hence if such a thing as control of the light flashes con-
sists In a muscular mechanism its action has no visible external
effects of contraction. So far as DuBois’ theory of control by
means of blood flow is concerned, we would have to place the same
limitations upon the mechanism, 1.e., action without any visible
external movements of parts, as in the theory of control by respira-
tion,for Du Boissuggests the dorso-ventral muscles as being at least
part of the mechanism. It might however, very easily be the
case that by the contraction of the heart, ete., we would be able
to get some sort of a rhythmical flow of blood through the organ
if the spaces did exist as he supposed. The results of the follow-
ing experiment show I believe, that both of these theories do not
account for the essential observed fact that the flashes are at irreg-
ular intervals, are sharply defined and that we must inevitably
admit the existence of a specific and direct control of the nervous
PHOTOGENIC ORGANS 447
system over the photogenic organ from a physiological standpoint.
The notes taken from only two out of several similar experiments
are reported here.
Experiment I. (a). I made a ventral incision in a fresh active
specimen on the segment anterior to the organ in order to cut the
nerve cord.
Effect: The organ remained dark. No spontaneous flashes
emitted though the animal was active being able to run and fly,
brush its antennae, ete. Mechanical stimulation caused a weak
light to appear at the point stimulated.
(b). A second specimen gave the same results except that a
very faint constant light appeared and remained. Shaking the
animal and otherwise stimulating it without stimulating the
organ directly gave no flash.
Experiment IT. (a). I laid open an active male Photuris
pennsylvanicus by cutting along the sides of the posterior three
segments, leaving the terga attached anterior and posterior.
Effect: The flashes were strong and normally controlled as
in fresh specimen.
(b). Then I raised and removed the connected tergae from
behind, also removed the viscera overlying the organ with a small
brush thereby exposing the whole dorsal surface of the organ
without injuring the nearby dorsal layer or anything laying
applied close to it.
Effect: Flashes normal, under complete control and not notice-
ably different from fresh specimen.
(ce). Then I made an incision anterior to the organ and close
to it
Effect: Light became extinguished except at points where
it apparently had been injured by manupilation and then only
showed weak and constant light at these points. Control abso-
lutely lost. Mechanical stimulation of the isolated organ caused
light to appear at point stimulated. This isolated organ when
placed in the tube and bulb apparatus responded to increases
and decreases in pressure by a marked increase in the intensity
of the light over nearly the whole organ.
In (ec) of Experiment 2 the blood flow if any was present must
have been interfered with. The dorso-ventral muscles were torn.
448 E. J. LUND
The terga carrying the spiracles and valves were removed, and
yet the flashes were normal and spontaneous and under perfect
control. It must be noted that the last nerve ganglion, nerves
and nerve chain were not dislocated in this experiment. Another
fact of almost equal importance is that the photogenic tissue,
apart from the nerve chain and other tissue, is highly irritable and
responds locally to stimuli. Where is the muscular mechanism
for such a response in this case?
Effects of temperature
Mangold (’10) has summarized the results of experiments by
others on the effects of different temperatures upon photogeny
(p. 345). No uniform results have been obtained among differ-
ent observers. But from those which have been obtained it is
clearly seen that there are optimum and maximum temperatures
for the process. I have found from a number of experiments
that with live animals or removed abdomens placed in a vial
into which a thermometer was fitted and the whole immersed
in water at 90° to 98° C. the temperature at which the light
was extinguished varied in the case of Photuris pennsylvanica
between 45° and 54° C. There was greatest intensity at about
40° C. Photinus sp. near maritimus gave results varying from
47° to 55° C. The specimens of this species were immersed with
forceps directly into water the temperature of which was registered
by a thermometer. With this species the temperature could be
raised considerably above the point at which light became extin-
guished and the light revived. This was noted again and again.
The maximum temperature from which the light was revived was
found to be about 84° C. The light reappearing at temperatures
around 50° C. In all cases where the whole animal was used it
died at about 40° to 45° C. and all control of the light was lost,
so that in P. sp. near maritimus the light was always continu-
ous after the specimen had been heated above a temperature of
40° to 45° C. In raising the temperature the light in P. sp. near
maritimus was always found to pass from the characteristic
greenish yellow to a yellowish, thence to yellowish orange and
finally to orange and sometimes into a reddish orange color before
PHOTOGENIC ORGANS 449
it became extinguished, the rapidity of these changes depending
upon the temperature of the water. The higher the temperature
the more rapid the changes and hence at high temperatures the
color would seem to change directly into orange before the light
became extinguished. At lower temperatures of the water (50°
to 60° C.) the transition in colors was gradual. Traces of this
same change in color was noted in Photuris pennsylvanica and
Photinus ebriosus. When the organs were cooled and the light
reappeared (P. sp. near maritimus) the transition in color of the
light was in the reverse order, i.¢., orange to yellowish and thence
to greenish yellow. This sequence in color changes becomes more
interesting when we compare them with color changes reported
to occur in certain marine photogenic organisms.
These experiments upon the live animal and the severed abdo-
men are obviously ill adapted to tell us definitely at what temper-
ature the particular chemical process between the formed sub-
stances which results in light takes place, for we have to take
into account the control and subsequent loss of control of the
organ, also the removal of the necessary supply of air or oxygen
to the organs and the formed substances concerned, upon the
death of the animal. There may also be other effects of temper-
ature which have nothing to do with the essential process of pho-
togeny itself but which indirectly determine whether it shall or
shall not take place. Yet certain facts stand out: (1), that the
process of photogeny is prevented from taking place at some temper-
ature; (2), that it may be continued if the material is not heated
above a critical temperature which prevents the possibility of further
action when the temperature is afterward lowered.
With the purpose of finding if possible whether the essential
chemical process concerned in photogeny is affected by certain
definite temperatures I performed a series of experiments upon
the luminous secretion of Cypridina squamosa (sp.?) and Cyclo-
pina gracilis’, two luminous salt water ostracods found upon the
bottom in shallow water at Montego Bay. The solutions were
made with pure filtered rain water, no distilled water being avail-
able at the time. Solutions with sea water were also tried in
7 Miss Mary J. Rathbun has had these and some other forms identified forme
and I wish here to express my thanks.
450 E. J. LUND
other experiments. No noticeable differences in the results
was found.’ Three or four of the ostracods were squeezed with
a pair of forceps in the test-tube containing the rain water,
thus causing in many cases a copious, Intensely luminous greenish
yellow secretion to be freed in the water. When this was shaken
a homogeneous luminous solution was obtained. The thermom-
eter was hung in the luminous solution in the test-tube and the
latter was immersed in a beaker of water kept at from 90° to98° C.
The alcohol lamp was enclosed in a hood so that no light from this
source disturbed the observations. The following are the tabu-
lated notes on the results of one series of experiments upon the
luminous secretion in a solution of rain water after the tempera-
ture effects had been noted in a general way by a preceding series
of trials.
SOLUTION IN
IMMERSION FILTERED RAIN RAISED TO RESULT
WATER
degrees
A 60 C. Recovered to a bright lumin-
escence upon cooling
first B 65 Recovered to a bright lumin-
Same solution { escence upon cooling
|second B’ 65 Showed only a very shght
recovery of luminescence
upon cooling
first C 67 Recovered to a less degree
Same solution 4 | than solution B’ at 65° C.
second | (CH 67 | Did not recover
Sane nition first D | 70 | Recovered very faintly
second IDY 7 Did not recover
E 71 Did not recover
F 70 Did not recover until after one
minute and then very faintly
G 71 Did not recover
Solutions raised to temper-
atures above 70° to 71° C.
never recovered their pho-
togenic power.
8 Thus there would seem to be no inherent reason so far as the essentials of
the chemical process are concerned why photogenic organisms should not be found
in fresh water, asis the case. Perhaps the photogenic function is to be considered
a primitive one so far as phylogeny is concerned.
PHOTOGENIC ORGANS 451
In all these experiments light disappeared at 50° C. and light
reappeared again at 50° C. The latter was determined with
solutions raised to 51°, 52° and 53° C., and then allowed to
cool, for when the solutions were raised to 60° or 65° C. and higher,
the temperature at which visible reappearance of the light took
place was not so well defined and sharp, because of the low inten-
sity and hence it was difficult to determine the exact temperature
at which the process began again. The time rate of return of
the light depends upon the time it takes for a solution to be low-
ered to the temperature where light reappears, for quick cooling
caused a quick return of the light.° Thus we see that here we
have a luminous secretion which when isolated yields definite
results while with the photogenic organs of the Lampyridae it is
more difficult to determine the exact temperatures, very probably
for the reasons given above. The important thing is that they
show exactly similar behavior toward temperatures.
Since the specific effect of temperature is one of the chief criteria
for showing the existence of enzymes these results have led me
to study further the osmic acid reaction in the tracheal end cells
of the Lampyridae. If the reduction of osmic acid on the tra-
cheoles and in the tracheal end cells depends upon the existence
of a specific enzyme then when the tissues of the organ are heated
above the temperature at which the enzyme is destroyed and
then treated with osmic acid we should presumably obtain no
reduction of the acid in the structures where the enzyme is located.
If this is the case we shall have found an explanation for the reduc-
tion of osmic acid in these cells.
The following are the summaries of results from six experiments
to test the effect of temperature upon the photogenic tissue with
regard to subsequent reduction of osmie acid in the tracheoles
and the tracheal end cells. All the experiments except Ex-
periment 5 had controls; the latter passed through the same
° In these experiments where intensity of light was noted the results depend
upon the visual sense of the observer. No apparatus for measuring the intensity
and point of disappearance of the light was at hand. But precautions were taken
to make all the observations in a perfectly dark night and against a black back-
ground.
452 E. J. LUND
processes except that of being heated. Sections were cut and
mounted, some stained and others unstained.
Experiment 1. Sixteen specimens of Photinus sp. near mari-
timus were taken.
(a). Control. From eight specimens I removed the abdomen
and placed them directly into 0.5 per cent osmic acid.
(b). Eight whole specimens were immersed in water at 98°
C. for one minute. The abdomens then cut off and placed in a
0.5 per cent solution of (OsO,) osmie acid thirteen hours (same
length of time as in (a).)
Result: (a). Reduction took place either in whole or greater
part of organ in all the eight specimens.
(b). No reduction of OsO, took place in any of the tracheoles
of the boiled specimens.
Experiment 2. Twelve specimens of P. ardens var.? were
taken.
(a). Control. Abdomens of six placed in 0.5 per cent OsOx;
three of these were cut and mounted.
(b). Six were placed in water at 99° C. in a dry vial for about
one minute. Then abdomens cut off and placed in the OsO,
solution.
Result: (a). In the three that were cut the tracheoles showed
strong reduction.
(b). No trace of reduction of OsO,; in any of the tracheoles
of the six specimens.
Experiment 3. Four specimens of Photuris pennsylvanica were
taken.
(a). Control. Abdomens of two immersed in OsQ,.
(b). Two immersed in boiling water one minute; abdomen
cut off and placed in 0.5 per cent OsQO, thirty-six hours (same as
(a)).
Result: (a). Strong reduction in tracheoles and tracheal end
cells.
(b). No reduction in tracheoles and tracheal end cells.
Experiment 4. Twelve specimens of P. sp. near maritimus were
taken.
PHOTOGENIC ORGANS 453
(a). Control. Six were placed in OsO, directly and left in
same time as (b).
(b). Six were placed in water at 80° C. for three to four min-
utes then into OsO,.
Result: (a). Showed strong reduction in most of the speci-
mens, and slightly less in one or two. All showed reduction.
(b). Three showed no reduction in tracheoles, while three
showed traces of reduction.
Experiment 5. Two specimens of P. pennsylvanica were im-
mersed in water at 714° C. until light became yellow, then orange
and was finally extinguished. The light did not return upon
cooling. The abdomens were removed and placed in the OsO,
solution. No control.
Result: Strong reduction took place in the tracheoles and
tracheal end cells.
Experiment 6. Twelve specimens of P. sp. near maritimus
were taken.
(a). Control. Abdomens of six placed in OsO,.
(b). Six placed in water at 60° C. (several minutes). Some
regenerated light slightly after immersing in the OsO;. All
passed through yellow then orange when immersed in the water
at 60°.
Result: Reduction took place more or less in tracheoles of all
specimens in both (a) and (b).!°
From a study of the results of these experiments with temper-
ature upon the intact organ, the luminous secretion and the effect
of temperature upon the osmic acid reaction, it may at once be
asked; is the reduction of osmie acid in the tracheal end cells
‘0 Bongardt (’03, p. 11) states that he succeeded in staining the tracheoles black
by ‘‘fixing the tissue in alcohol or sublimate-acetic, staining in borax carmine,
decolorizing in acid alcohol washing in water twenty-four hours. placing in solu-
tion of 1:400 OsO, over night then into strong acetic eight to ten hours, washed in
distilled water, hardened in alcohol embedded and cut.’ Another method is also
mentioned using gold chloride, which is more complex. From such treatment
of the tissue it is obviously difficult to learn anything about the nature and where-
fore of the reduction, which in such a method may be due to entirely different
causes.
tHE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, No. 4
454 E. J. LUND
and tracheoles dependent upon the vitality of the protoplasm of
the tracheal end cell or is it only dependent upon the existence
of the formed substances which by themselves, as has been shown
by drying the organs, are able to produce the phenomenon of
photogenesis? With this comes the question, did the tempera-
tures below that at which osmie acid reduction is prevented from
taking place kill the protoplasm of the photogenic cells? An
answer to the latter will furnish an answer to the former question.
The animals are very sensitive to higher temperatures and dry-
ness. In all the temperature experiments upon the whole animal
the specimen always died at about 45° to 50° C. if left in for as
long a time as one minute. Control of organ was lost though
very often a residual glow, sometimes quite strong, remained
after cooling if it had not been heated to temperatures which per-
manently prevented the continuance of the light. It will be noted
that strong reduction took place even as high as 713° C. and that
traces of reduction were evident in Experiment 4, where the
organs had been heated to 80° C. for three to four minutes. Fur-
ther it is important to note that the reduction of the acid does
not depend upon the continuance of the process of photogenesis
while the acid is penetrating into the photogenic layer as is plainly
shown in the reported experiments (4 and 5). Therefore all the
conditions which are necessary for the process of photogenesis
are not necessary for the osmie acid reduction in the tracheal end
cell, it is only necessary that the organs be not heated above a
certain temperature.
It is of course impossible to prove that the protoplasm of the
cells of the photogenic organ was killed at 713° or 80° C. as in
Experiment 4. Another important fact to be remembered how-
ever is that osmic acid is itself a strong killing and fixing agent.
The longer the time that the organs are left in the solution the
more extensive (within certain limits) is the osmic acid reduction
in the tracheal end cell (fig. 2) now if a small amount of reduction
has taken place around the fureation then evidently we must have
had the penetration of the acid to the tracheal end cell and this
it seems would be efficient in destroying its vitality. If then
the first traces of the acid are sufficient and the time long enough
PHOTOGENIC ORGANS 455
for the killing of the cytoplasm then the further reduction of the
osmie acid can not have been due to the vitality of the cyto-
plasm but to some substance in the cytoplasm which by itself
is able to reduce osmic acid. The temperatures at which the
possibility of return to the conditions necessary for photogenesis
becomes permanently destroyed in the Lampyridae, viz.. 70° to
84° C., corresponds in general to the temperature which destroys
the possibility of the osmic acid reduction in the tracheal end cell
to take place. When tincture of guaiacum is applied to the tissue
of the photogenic layer it does not turn blue, neither does it turn
blue when hydrogen peroxide is added though the hydrogen perox-
ide is decomposed. The power of decomposition of hydrogen
peroxide is not limited to the photogenic organs but any of the
body parts when placed in this reagent decompose it showing the
presence of a ‘catalase.’ I immersed specimens for one minute
in boiling water and tested parts of the body with hydrogen
peroxide. No decomposition took place showing that the ‘cata-
lase’ was destroyed. In view of these reactions of osmic acid
we may consider it certain that the normal reduction in fresh
photogenic tissue is due to the presence of some substance prob-
ably of the nature of a reductase which is formed in the cytoplasm
of the tracheal end cell and peritracheolar membrane.
USE OF THE ORGANS
Apart from the problems concerning the chemical-physiologica]
mechanism of the luminous organs of the Lampyridae, Elateridae
and numerous other forms, is the significance and ultimate use
of these organs to their possessors.
Osten-Sacken (’61) states that he found males of Photinus
pyralis while themselves emitting flashes of light, to respond posi-
tively to the light from the females, and that a short time there-
after he found them copulating. Emery (’86) also found the
males to respond to the females in the same manner though
he did not observe a subsequent copulation.!: It is a matter of
' Recently MeDermott writes me that he ‘has very definitely confirmed”
Osten-Sacken’s observations for Photinus pyralis, and extended them to P. con-
sanguineus and P. scintillans.
456 E. J. LUND
common knowledge to people of regions of the tropics where
Pyrophorus abounds, that it responds under certain conditions
to ordinary artificial light stimuli. I have collected P. plagio-
phthalmus by means of a small partly enclosed incandescent light
and by means of it have found them to react strongly at distances
of fifty yards or more. During this reaction all their photogenic
organs are in full activity.
Comparison of the eyes and the photogenic organs in those
species of Lampyridae where both males and females were ob-
tained, showed. (a) That in all males, and to a smaller extent
in those species, the females of which are most active and abun-
dant, the eyes are greatly developed. (b) That in all cases the
eyes and at the same time the photogenic organs of the males are
larger than the corresponding organs of the female, i.e., the extent
of development of the eyes is in direct relation to the extent of
development of the photogenic organs and activity. In one large
female (unidentified) which I found on the ground among some
grass the organ was exceedingly small, the eyes were very much
reduced in size and added to the above shortcomings the wings
were rudimentary. The photogenic organ, however was very
active and well controlled giving off a strong greenish yellow light.
This becomes of further interest when we note that there exists
a direct relation between the head ganglia and the control of the
photogenic organs. When the head is removed from specimens
control of the light is interfered with or lost and the spontaniety
of the flashes is also lost. Removed abdomens which contain
the last abdominal ganglion show no spontaneous control, usually
the organs give off a continuous or irregular glow or else light dis-
appears though they respond to mechanical and some other kinds
of stimuli.
A syllid, probably Odontosyllis pachydonta Verril, during
certain periods, reacted to each other and to an artificial light
stimulus in a most striking manner. The eyes of this syllid are
unusually developed. In one ease a female—one of the sylliidae—
filled with eggs was taken while brillantly luminous, and some of
the eggs, which were shed readily, were fertilized with sperm
taken from individuals caught in the tow net in the same place
PHOTOGENIC ORGANS 457
and at the same time. The next morning these had developed
into late cleavage stages. Galloway and Welch (711) have found
by direct observation that the positive response of the male and
female of Odontosyllis oenopla to each other results in the bring-
ing together of the eggs and sperm in the water so that here we
have one of the first well established instances of the sexual adjunct
significance of photogenicity in organisms.
SUMMARY
1. No fundamental structural difference exists in the photo-
genic organs of the species of Lampyridae studied. The elemen-
tary photogenic mechanisms of which the organs are made up,
viz., the trachea, tracheoles, tracheal end and photogenic cell
with their relation to the nervous system, taken collectively, are
the same in all the forms.
2. The tracheoles are tubular and have been shown to anas-
tomose in Photinus ebriosus. They are made up of a chitinous
substance, which is more resistant than the protoplasm to re-
agents. Their number, where it could be definitely determined
has been found to be constant for each tracheal end cell. The
tracheoles are in no case limited in their course to the outside of
the photogenic cells but penetrate into the cytoplasm of the latter.
They are apparently filled with a liquid under such conditions
as make observation of them possible.
3. No distinet cytoplasmic membrane around the tracheoles
could be demonstrated yet the reactions to osmic acid shows that
a limited region about the tracheoles has the same specific prop-
erty of reducing the acid as the cytoplasm of the tracheal end cell.
4. The reduction of osmic acid upon the tracheoles and in
the cytoplasm of the tracheal end cell is dependent upon the pres-
ence of a substance probably of the nature of a reductase which
shows the same properties as regards the effects of temperatures
upon it as enzymes in general. The process of photogenesis is
dependent upon the presence of this substance as is shown by the
parallel effect of temperature upon the production of light.
458 EB. J. LUND
5. The process of photogenesis is independent of the vitality
of the cytoplasm and is a resultant of the interactions of formed
substances in the presence of water and oxygen. It is highly
probable that it partakes of the nature of an oxidation but this
has yet not been demonstrated for the photogenic process in the
Lampyridae if in any animal.
6. It has been shown that photogenesis is incident upon the
utilization of a nitrogenous compound—the photogenic granules—
giving staining reactions like those of lecithin and different from
those of the true fats, and that this nitrogenous compound appears
at least in part at the endof the process in the form of a nitrogenous
waste product. This crystalline substance appears from its re-
actions to be allied to or identical with some of the split products
of nucleic acid.
7. The dorsal layer cells become the repositories for the waste
product. No direct transformation of the photogenic cells, as
such, into cells of the dorsal layer takes place.
8. The photogenic process is localized in and adjacent to the
cytoplasm of the photogenie cells and especially (as far as could
be determined) where the cytoplasm of the tracheal end cell, and
tracheole is applied to the photogenic cell.
9. The increase in intensity of the light resulting from increase
in pressure is due to the increased oxygen content in the regions
where photogenesis takes place. Changes in the oxygen content
is not the primary means of control of the organs.
10. The primary control of the organ is by nerves in direct
connection with the photogenic tissue and not by an external
respiratory muscular mechanism. The termination of nerve
fibers in the tracheal end cells as Bongardt states he found in
Lampyris splendidula is supported by the fact that photogenesis
may be limited to points which can structurally only be referred
to the tracheal end cells and that upon increased stimulation a
phenomenon similar to a ‘spread’ of the stimulus takes place.
Furthermore the photogenic tissue is irritable and responds locally
to mechanical stimuli.
11. A direct control relation exists between the photogenic
organs and the nerve centres of the head. This is apparently
PHOTOGENIC ORGANS 459
correlated to the relation which exists between the degree of
development of theeyes and the photogenic organs, 1.e., the extent
of development of the eyes is in direct proportion to the extent
of development of the photogenic organs.
12. The positive response to light by some photogenic organ-
isms results in bringing the eggs and sperm in nearer proximity
to each other in cases where these are set free in the water before
fertilization. This has been shown for Odontosyllis enopla by
Galloway and Welch. This is also very probably true for Odon-
tosyllis pachydonta (?) Verrill. In other cases the response may
lead to copulation as has been found for Photinus pyralis by
Osten-Sacken, and for Photinus consanguineus and P. scintillans
by McDermott.
Zoological Laboratory
Johns Hopkins University,
September 20, 1911.
LITERATURE CITED
Bonaarpt, J. 1903 Beitrage zur Kenntniss der Leuchtorgane einheimisher
Lampyriden. Zeits. f. wiss Zool., Bd. 75, pp. 1-45.
Crark, E. D. 1910 The plant oxidases. Dissertation, Columbia University.
Dusors, R. 1886 Contribution a l’étude de la production de la lumiére par les
étres, Les Elatérides luminaux. Bull. Soc. Zool. France, Année I,
pp. 1-275.
1892 Sur le méchanisme de la production de la lumiére chez l’Orya
barbarica de Algerie. Comptes Rend. Acad. Sci. France, 1892, exvii,
pp. 184-6.
1895 Physiological Light. Smiths. Inst., Wash., D. C. Report for
1895, pp. 413-431.
Ermer, T. 1872 Bemerkungen iiber die Leuchtorgane der Lampyris splendidula.
Archiy f. mikr. Anat., Bd. 8 p. 653.
Emery, C. 1884 Untersuchungen iiber Luciola italica. Zeits. f. wiss. Zool.,
Bd. 40, pp. 338-355.
1885 La luce della Luciola italica, osservata col Microscopio. Bull.
Soc. Entomol. Ital. Ann. 17, Trim 3/4, p.351-5. Also Jour. Roy. Mier.
Soc., 1886 p. 234.
1886 La luce negli amori delle Luciole. Bull. Soc. Entomol. Ital.,
Ann. 18, p. 406.
460 E. J. LUND
Emery, C. 1886 Sur le lumiére des Lucioles (Luciola italica). Arch. Se. Phys.
et Nat. (Genéve) (3), T. 14, no. 9, p. 272-5.
Gatitoway, T. W., and Weicu, P. 8. 1911 Studies on a phosphorescent Bermu-
dan annelid, Odontosyllis enopla Verrill. Trans. Am. Mier. Soc.,
January 1911, vol. 30, pp. 13.
Heineman 1872 Uber die Leuchtorgane der in Vera Cruz vorkommenden
Leuchtkifer. Archiv. f. mikr. Anat., Bd. 8, pp. 461-471.
1886 Zur Anatomie und Physiologie der Leuchtorgane Mexikanischer
Cucujos (Pyrophorus). Archiv. f. mikr. Anat., Bd. 27, pp. 296-383.
HoitmGren 1895 Die trachealen Endverzweigungen bei den Spinndriisen der
Lepidopterenlarven. Anat. Anz., Bd. 11, pp. 340-346.
Ives, E. H., and Cospitentrz, W. W. 1910 Luminous efficiency of the firefly’
Bulletin of the Bureau of Standards, Wash., D. C. vol. 6, pp. 321-336.
IKxastie, J. H. 1910 The oxidases and other oxygen catalysts concerned in bio-
logical oxidations. Hygienic Laboratory, Pub. Health and Mar. Hosp.
Serv. U. 8., Bull. No. 59.
Koéturker, A. 1857 Uber das Leuchten der Lampyris. Verhandl. d. Wiirzb.
phys-med. Gesell., 1857. Bd. 8, pp. 217-224.
isc4 Uber den Bau der Leuchtorgane der Ménnchen der Lampyris
splendidula. Sitzungsber d. Niederrh. Gesells. f. Nat. und Heilkunde.
Le Bon, G. 1908 The evolution of forces.
Liuuig, R.S. 1902 Oxidative properties of the cell nucleus. Am. Jour. Physiol.,
vol. 7, p. 412.
LoiseL, M. G. 1903 Essai sur la technique microchemique comparative de la
lecithine et des graisses neutres. C. R. de la Société de Biologie,
vol. 55, p. 703.
Lunp, E. J. 1911 Notes on light reactions in certain luminous organisms.
Johns Hopkins University Circular, Feb.
Macartney 1810 Observations upon luminous animals. Phil. Trans. vol.
100, pp. 277-279.
Mancoup, E. 1910 Die Production von Licht. Handbuch der vergleich. Phys-
iol. (Winterstein), pp. 225-392.
Mann, G. 1901 Physiological Histology, Oxford.
Mouuisu, H. 1905 Luminosity in plants. Smiths. Rep. 1905, pp. 351-362.
Nurrine, C. C. 1899 The utility of phosphorescence in deep sea animals. Am,
Nat., vol. 33, pp. 792-799.
1910 The theory of abyssal light. Proce. 7, Cong. Zool.
Osten-Sackren, Baron. 1861 Die amerikanischen Leuchtkafer. Stettiner
Entom. Zeitung, vol. 22, pp. 54-55.
Prerers, A. 1905 Luminescence in etenophores. Jour. Exp. Zodl., vol. 2.
PHOTOGENIC ORGANS 461
Rippue, O. 1910 Studies with Sudan III in metabolism and inheritance.
Jour.
Hxp. Zodl., vol. 8, p. 163.
ScuHrerner, O. and Suttivan, M. X. 1911 Concurrent oxidation and reduction
by roots. Bot. Gaz., vol. 51, p. 273; also the same journal, vol. 51, p. 121.
Scuuttzp, M. 1865 Zur Kenntniss der Leuchtorgane von Lampyris splen-
didula. Archiv. f. mikr. Anat., Bd. I, p. 124.
Starke, J. 1895 Uber Fettgranula und eine neue Eigenschaft des Osmium-
tetroxydes. Du Bois Raymond. Archiv. f. Phys., p. 70.
TowNsHEND, A. B. 1904 The histology of the light organs of Photinus margin-
ellus. Am. Nat., vol. 38, pp. 127-148.
Warasn, S. 1895 Physical basis of animal phosphorescence. Biol. Lects.,
Woods Hole.
1898 Animal luminosity. Biol. Lectures, Woods Hole.
WeirtLaNner, F. 1909 Etwas vom Johanneskaéferchen. Verh. Zool. Bot.
Wein, Bd. 59, pp. 94-108.
Ges.
WieLtowigsski, H. R. von 1882 Zur Kenntniss der Leuchtorgane von Lampy-
riden. Zeits. f. wiss. Zool., Bd. 37, pp. 354428.
1889 Beitrage zur Kenntniss der Leuchtorgane der Insekten. Zool.
Anzeig., Jahrg. 12, pp. 594-600.
WISTINGHAUSEN
1890 Uber Tracheenendigungen in den Serikterien der Raupen.
Zeits. f. wiss. Zool., Bd. 49, p. 565.
PLATE 1
EXPLANATION OF FIGURES
3. Photomicrograph of part of a longitudinal vertical section of photogenic
organ of o& Photuris pennsylvanica, killed in 0.5 per cent OsO,, thirty-six hours.
Weakly stained in Ranvier’s picro carmine; high power; D, dorsal layer showing
cell walls and content of granular waste product; P, photogenic layer showing
nuclei NV. and photogenic granules G.; V, vertical tracheae; F, furcations where
osmic acid reduction first takes place; two tracheoles arise from each; S, sternite,
Y, tracheal end cells between dorsal and photogenic layers from which tracheoles
pass ventrally; similar ones may be found scattered over the ventral surface.
4 Horizontal section through photogenic layer of o& Photuris pennsylvanicus;
Lyons blue and borax carmine; L, lumen of vertical tracheae; NV, nuclei of tracheal
epithelium; £, nuclei of tracheal end cells forming a cylinder around vertical
trachea; P, nuclei of photogenic cells. This photograph shows the so-called ‘cell
boundaries’ and grouping of the photogenic granules about the nuclei.
PHOTOGENIC ORGANS PLATE 1
E. J. LUND
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, No. 4
463
PLATE-2
EXPLANATION OF FIGURES
5 Photomicrograph of a vertical longitudinal section of photogenic organs of
Photinus sp. near maritimus, taken with the aid of a ‘bulls-eye’ condenser by
reflected light; the posterior and part of anterior photogenic organ is shown, with
the relative distribution of the crystalline nitrogenous waste product. A, among
the abdominal vicera; D, dorsal layer cells completely filled; P, periphery of pho-
togenic cell; NV, deposit around the rows of nuclei of photogenic cells. Note rela-
tive thickness of dorsal and photogenic layers in this condition of the organ.
6 Photomicrograph of high power of photogenic and part of dorsal layer
showing distribution of crystalline deposit. D, dorsal layer, cell outlines and posi-
tion of nuclei barely indicated; P, photogenic cells greatly diminished in size and
containing the waste products in the peripheral region of the photogenic cell and
around the nuclei (cf. figs. 1, 2 and 5).
464
PLATE 2
PHOTOGENIC ORGANS
E. J. LUND
6
THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 11, NO. 4
165
PLATE 3
EXPLANATION OF FIGURES
7 Camera lucida drawing of part of vertical section of photogenic organ of
Photinus sp. (?) (Jamaican); fixed in boiling water, stained in phospho-molybdie
acid haematoxylin. #, tracheal end cells forming eylinder—shown in section—
around vertical trachea, V which terminates in branches on ventral side; 7’, tra-
cheal trunk showing hairs;
Z, peripheral zone of photogenic cell; P, photogenic
granules V. nuclei of tracheal epithelium; D, dorsal layer cells; HW, hypodermis;
C, chitin of sternite bearing hairs.
8 Camera lucida drawing of part of cross section of photogenic organ of
Photuris pennsylvanica; fixed in osmiec acid. The photogenic layer is one cell
deep and the tracheal end cells are distributed over the dorsal and ventral sides
instead of forming cylinders around the tracheae. 7, trachea from which
branches B, arise passing into the two partly embedded tracheal end cells C;
n, nuclei of tracheal e.c., F. fureation from which arise three tracheoles. Reduc-
tion of the acid has not been sufficient to show the tracheoles throughout their
entire course. A thick deposit of the reduced OsO, is shown about the fureation
with traces of it in the cytoplasm, nucleus and region between the photogenic and
tracheal end cell; H, nuclei of hypodermis. J, chitin; D, dorsal layer cells, note
relative thickness of dorsal and photogenic layers; P, photogenic cells with gran-
ules.
9 View of portion of ventral surface of photogenic organ of Photinus sp. near
maritimus, drawn by means of cam. luc. outline of active organ. SS, shaded area,
corresponding to eylinder and showing only a faint light due to diffusion of hght
from photogenic areas L. The latter correspond to region of photogenic cells. P,
peripheral region of photogenic cells showing under some conditions a slightly
greater intensity of light; V, nuclei of photogenic cells faintly visible as small
shaded spots.
466
PHOTOGENIC ORGANS PLATE 3
E. J. LUND
THE JOURNAL OF EXPERIMENTAL zOOLOGY, vou. 11, No. 4
467
EXPERIMENTS ON DEVELOPING CHICKENS’S EGGS
STEWART PATON
Biological Laboratory, Princeton University
The results of observations made upon the developing eggs of
several species of selachians suggested the possibility of repeating
these experiments, with certain modifications, upon the embryos
of other vertebrates.
The egg of scyllium canicula is admirably adapted for studying
the primitive movements of the heart and myotom without in
any way disturbing the normal relationship of the growing organ-
ism, but on account of the difficulty of securing these eggs in
large quantities a search was made for material which could be
more easily obtained. Observations made upon several species
of lizards, frogs and fresh water fish were for many reasons unsatis-
factory and attention was then directed to the chicken’s egg.
Experimenters have at various times removed the fertilized
egg from the shell, and after detaching the embryo, have suc-
ceeded in keeping the latter alive for some time not, however,
exceeding a period of twelve hours. After many futile attempts,
an Operative technique has been devised making it possible in
the majority of instances to remove the fertilized chicken’s egg
from the shell, place it in a glass dish containing fluid and return
the receptacle to the incubator, when development under the
conditions to be mentioned proceeds uninterruptedly. The egg
freed from the shell becomes an object for observation and exper-
iment, and not only the incidence of the primitive movements
of the heart, but also many other interesting phenomena con-
nected with the growth of the embryo may be observed and
recorded. The technique employed in the operations is as
follows:
THE JOURNAL OF RXPERIMENTAL ZOOLOGY, vou. ll, No. 4
469
470) STEWART PATON
All solutions are sterilized in the autoclave. Such a small
quantity of fluid is lost during the process that in the majority
of cases it is not often necessary to replace it but, if in certain
cases it is essential, this may easily be accomplished if the fluids
have been sterilized in graduated flasks. In order to shorten
the operation as much as possible, and to minimize the risk of
exposing the sterilized fluids to the air, the solutions are poured
into dishes in which the eggs are to be placed, covered and put
in the thermostat. The lids, which should be 5 mm. to 10 mm.
larger than the dishes, rest upon collars of cotton held in place by
string, and by this means free access is given to the air. Care
should be taken that the cotton does not come into contact with
the fluid in the dishes, and on the other hand, these collars must
be sufficiently thick to raise the lids and give plenty of opportunity
for the passage of air. Many embryos are killed by a deficient
supply of oxygen. The cotton acts as a filter and prevents all
bacteria except those within the shell from contaminating the
fluids in the dishes.
After the egg has been for the requisite amount of time in
the incubator, it is removed, the shell is wiped off with 95 per cent
or preferably 100 per cent alcohol, and with the aid of a pair of
forceps that have been sterilized, an opening with smooth edges
is made in it and the contents allowed to slide gently into the
dish containing the fluid which should be of the same temperature
as the egg.
If the dish contains sufficient fluid the egg will quickly right
itself so that the embryo is on top. Even slight differences of
temperature seem to be fatal to the success of the experiment,
and on this account, it is better to conduct the whole process of
transferring the egg from its shell to the dish in some kind of
warm chamber, such an one as can readily be constructed in the
laboratory. When the egg is in the dish and covered, the proc-
ess of development may be observed through the glass top with-
out exposing the contents to the air.
The earlier in development that the transfer is made, the greater
is the chance of failure, but when the embryo has attained the
size seen under normal conditions at about the 26th-27th hour
EXPERIMENTS ON DEVELOPING CHICKENS’S EGGS A471
of incubation, the operation is nearly always accomplished with-
out serious injury to the growing organism.
The action of a variety of fluids upon the embryo were observed
and a brief account of some of the effects that were noted will
now be given:
The constituents of Ringer’s solution were tried singly and in
combination. NaCl in varying strength from 0.5 per cent to
2 per cent if uncombined, with other salts at once killed the
embryo, but development although apparently taking place
at a slower rate than normal, followed when the egg was placed
in 0.7 per cent solution of NaCl to which 2.7 cc. of a molecular
CaCl, solution was added. The record of all my experiments
show that the rate of development is retarded in the sodium-
calcium solutions, and the vitality of the embryo is also weakened.
The extremely interesting fact in connection with this experi-
ment is that the presence of such minute quantities of calcium
is sufficient not only to protect the life of the embryo, but also
to insure at the proper time the incidence of the cardiac move-
ments. The same observation with practically similar results
was made upon the eggs of trout. In the presence of these minute
quantities of calcium, when combined with sodium, the regular
and rhythmical pulsations of the heart begin, and are continued
for several hours with increasing force and rate, but later the
embryo dies.
The calcium alone is not sufficient to insure the continuation
of the developmental processes. In solutions to which KCl
(6.3 ec. of a molecular solution) has been added, in addition to
NaCl and CaC, growth seems to take place at a normal rate.
The effect of MgCl, alone upon the growth of the embryo chick
and its relations to the primitive movements of the heart has not
been experimentally determined, but my impression is that the
solutions containing MgCl, when not combined with CaCl, and
NaCl are highly toxic.
The effect of urea in strengths varying from 0.5 to 2 per cent
either alone or combined with NaCl does not act as a stimulus
to growth and the embryo soon dies when placed in solutions
containing this substance.
472 STEWART PATON
Embryos detached from the egg and floated in any of the solu-
tions named, live but a short time and the incidence of the prim-
itive movements of the heart in these detached specimens is
never observed.
Caution should be observed in basing any deductions in regard
to the immediate action of the various salts upon the incidence
of the cardiac pulsations. It does not seem probable that the
failure of the detached embryos to develop is purely the result
of shock, but is due chiefly to the absence of nourishment supplied
by the egg.
BINDIXC
==_ 7, JUN 2 2 1966
QL The Journal of experimental
1 zoology
J68
Viel!
cop.2 \A\\
Biclogical
& Medical
Serials
PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET
UNIVERSITY OF TORONTO LIBRARY
Eye pant
ht
ere tha
sua
>
2
SS
=
SEES
nes:
mee
oy
ie ;