THE COIIIIELATION OF SOJIATIC CHARACTERS, ETC.
‘JIit1
107
(‘orrelation of Yom:itic C‘haracters a i d Chromatin Rod-Lengths, beiiig
ii Further Stucly of (:Iiroiiiosonic Uiiiieiisions.
By (.’. E’. U. MEEK,
JI.Sc., F.L.S., F.Z.S.
( W i t h 5 Test-tigiires.)
[ltead 20th
Julie,
1912.1
ISTItOUUCTIOS.
I N a. recent paper * dealing with cLLroiiio.~onie cliiiieiisioiis in iiuiiierous
orgiiiisiiis, I Iiave sliowi~t1i:it throughout tlic aiiiiii:d kingdom lengths of
cuitiponriit rod9 :tpl)ear to constitute ineiiiLers of :i geiieritl series in iirithiiietical l)rogression, wliereiis only t.liree diaiiieters exist, viz. -21 p iii
P r o t o z o : ~ :iiicl
,
*42 and -83 p in low ;tiid liiglier Metaxon respectively.
( ‘oiisit1er:itioii of the results g i w n Iias, iiioreover, lctd to tlie eii~uiciationof
:in hypothesis wliicli 1)ostul:ites a ,wries of c!-cles in the coursc of pliglogcny.
Tt, is suggwtcd t h t tlie cliroiiiatiii griiiiiiles of the siniplest Protozoa have
1 ) ~ e iconverted
i
into rods by puroly 1inc:ir growtli, :~ccoiiiptiiiyingcvolutionary
dt!rc.Iopnicwt, : i i i t l iiicrewiiig soiiiittic c+oiiiplesity,: l i d , siiice the rate of this
growth cniiiiot 1i;ive been the s:inie in all cliroiiiosoiiies, rods of various lengths
1iiive I m i i evolved ; esaiiil)les of siich conipleses caii be st!eii iii (’ilicitu and
other liiglily cliEcrcnti:itcd Protozoa. A stage in phylogeny w a s later reached
w l i c i i ii iiiiisiiiiuiii rod-lengtli Iiad bcttn :ittaiiied, such limit liiiring been
iiiili~sed by spiiidlc iiiec1i:iiiisni or othcr 1)liysical coiiditioiis ; when this
occuriwl c1ironi:itiii units con jugatecl in fours, :uid tlie iior1iial thread-width
\ w s thus doul~led. The chroiiiosoiiies, rcducecl in number, tlieii segmented
tixiisveiwly iiit,o iiuiiierous spheres of tlie i i e w diameter, and the process,
wliicli :ipprosiinately re-est:iblislied tlie I i u i i i l m of cliroiuosonics 1)reviously
wen, cii:ibletl thein to ciitcr a f i w h coiirst’ of lincnr growth :iccoiiipaiiying
furtlier cvo1ut)ionary dcvelopiiieiit. In this iiiniiiier tile coniplescs of low
Al(1tiizo:i iiiay have wolvetl froiii tliow of 1’rotoxo:m uiiccstors.
\\‘lien tlie Ieiigth-limit of chroiiiosoiiies WIIS again reaclid, coiijugution of
iiiiits once iiiore occurred, arid this WIS followed as before by seginentatioii
into spheres of t h iiow diaiiietei* ; tlw last n:inied having twen thus doubled,
becatlie ideiitical with tliiit now foiiiicl in orgiiiisiiis beloogiiig to phyla ahove
a i i d includiiig Neiii:itlieliiiiiitliia.
This tlio chroiiiatiii t,hread-width of the
high l\let;izoa. 11i:tyhave evolved froni thot of t,he lower.
s e ( u s to ~iccordwith plieiioiii(wa, for I liavc been able to
find in tlicl :iniiii:il kiegdoiii examples that :ippireutly represent stages of
* ‘‘ .k Netricsl -1ntilysis of Clironioeome Complexcs,” l’hil. Trans. lloy. Soc. ser. 13,
vol. 303, 1912.
108
MR. C. F.
r.
MEEK OX THE CORRELh’l’ION O F
transition to a greater thread-width : it is, however, impossible to prove this
If it is
phylogenetic cycle with the meagrc data at present avai1:tble.
eventually cstablished, we must realize that an attempt to correlate bomatic
characters and individual cliromosoines must fail the moment that we conbider
any but tho most closely allied organisms: at definite periods a complete
rcarrangeinent of units has occurred, and, since the subsequent rate of growth
must have varied in different chromosonies under different conditions of
environment, we have no reason for assuming correspondence between rods
of the same length found in the germ-cells of widely separated organisms.
Within the limits of :L genus, howrver, it may be possible to trace somatic
differences to differences in chromatin growth, for closely allied animals must
have developed along the ~ 1 1 1 0or par:illel lines, and we may therefore be able
to identify corresponding chroniosomcs in their respective complexes. Jn the
paper already referred to I have given camcra lacidu. drawings of chromosome
rods composing the coiriplexes of several species of Stenotot7mrs, and have
shown that the lattrr can be individually distiuguislied by the presence or
absence of certain rod-lengths ; I now propose to deal with another species
of this genus in order to show that this plienonienon is prob:tbly common to
all its members. Moreover, the comparative study of allied species may
enable us to est:dilish some correlation with rcspect to length of cliromosomes
and somatic characters.
MATERIAL
AKD METHOLM.
Steiiobotlirzrs czrvtipennis, which belongs to tho tribe ‘l‘ryxali(1zc and the
family hcridiida, is not found in the British Isles, and I am indebted for
the material to the kindness of Prof. H. S. Davis, wlio sent me testes fixed in
Hermann’s solution and embedded in paraffin, from the University of Florida,
Gainesville, U.S.A. The sections were cut 8 p thick and stnincd with
Heidcnhain’s iron hamatoxylin, thc mordant used being an aqueons solution
of iron iiluni. The preparations were studied by means of a. Zeiss apochromatic oil-immersion objective of 2 min. focus and N.A. 1.30, in conjunction
with compensating oculars nos. 6, 12, and 18 : I have used throughout the
holoscopic oil-immersion substagc condenser made by Mcssrs. Watson, of
High Holborn, London,
All drawings were inade with the aid of :I large Abbe camera lucida a t a
magnification of three thousand and forty-eight diumcters, the magnification
heing estimated by means of a stage-micrometer graduatud to rcml oneliundretlth of a millimetrc. When necessary, resolution was facilitated by
interposing u. Gifford screen.
111 order to avoid error due to foreshortening, drawings have lieen made
only of‘ chroniosoincs that 1:iy at right angles to the microscopic line of vision
SORfATlC CHARACTERS AND CHROMATIN ROD-LENGTHS.
109
throughout tlieir entire length, and errors in draughtsmanship have been
minimised by drawing each individual chromosome many times ; the
measurements given should therefore represent the true dimensions with as
great accuracy as can be obtaiiied with the means now a t our disposal.
SPERMATOGENESIS.
The testes are two ovoid bodies lying in the middle of the abdomen
dorsally to the alimentary canal ; they :ire composed of tubular follicles
tapering towards the ends and divided into numerous tracts and cysts. The
primary and secondary spermatogonin lie at the extreme anterior end of each
follicle, and nest to these are large areas occupiecl by cells undergoing the
growth-period : no resting stage occurs between the mat.uration mitoses, and
the nest portion of the follicle is accordingly occupied by both prilnary and
secondary sl~erniatocyte divisions. The posterior end contains spermatids
undergoing tr:ineforinntion to unripe end ripe sper~n:itozoa. As I have
already pointed out in the case of S. v i ~ k h l t ! all
, ~ ,cells in one cyst are not a t
precisely the s:ime stage of development, :~nd in a transverse section mitotic
figures and resling stages can be seen Iyillg sitle Ly side.
Tbs primary slicrinntogonia :ire arranged i l l n cluster a t the extreme
ante Yior end of the follicle, lier re as the secondary spermatogonia, more
posteriorly placed, are in greater numbers
and appear to be without definit,e arrangement in the cyst. The chromatin during
the resting stages is disposed in granules
upon linin threads, and the nucleus is apparcnt>ly :I complete reticulum. I n the prophase oE division this network breaks into
numerous filaments, which shorten arid condense until seventeen compnct and smooth
clironiosomes are seen lying in the equatorial
planc ; thcsc are divisible into eight pairs
Fig. i . - ~ : p ~ r , , , ~ ~ t ~Illetaphase.
~~,,i~l
and one odd chromosome, which corresponds
\\.it11 the non no some of Davis and the heterotropic, accessory, and S chromosoine of other writcre.
Tlle sixteen ordinary cliroinosonies :ire graded in length and individually
composed of t\vo equal rods, of which one passes to each pole in the subsequent : t ~ ~ a ~ ~ ’:hthe
a s c plnne of cleavegc is invariably pwallel to the major
axes of tllesp rO,]S, \\y)lich appear to be indivisible units. The diameter of the
ordinary rotls i3 coiist:int, whereas that of the odd chromosome is greater and
varies throughont its lc.ngtl1 ; the latter is thus easily distinguishable from
110
MR. C. F. U. MEEK ON THE CORRELATION OF
the other members of the complex. Fig. 1 shows a polar view of this
metaphase, and seventeen chromosomes are seen in the equatorial plane.
The last spermatogonial division is followed by a long period of growth,
and large tr:icts of the follicle are seen occupied by cells undergoing this
stage.
The chroinatin is again disposed in
granules upon n network of linin threads, but
the odd chroinosoine takes no part in the general
dissociation and remains as a darkly staining and
homogeneous body apposed to the nuclear membrane. The prophase of the first niaturation
division is characterized by tlie fission of individual granules and the breaking of the network
into numerous double filainents ; these, which are
Fig. 2-The Growth Period. at first long and ragged, conjugate in pairs and
condense into the usual tetrads, appearing as rings,
crosses, and figures of eight. The closeness with which the componcnt rods
are folded upan one another makes resolution extremely difficult, but sizcrelationships corresponding with those seen on the sperniatogoriial spindles
are again recognizable, and each tetrad is undoubtedly composed of four
equal rods in juxtaposition. Since these rods are similar to those oE the
earlier mitoses, the total amount of chroinatin remains unchanged, and the
l?i"lgS.3
& 4.-Polar and lateral views of first maturation mitosis.
eight tetrads are collectively equivalent to the sixteen ordinary sperinatogonial chromosomes. I have failed to determine whether this division is
reductional or equational, but this is immaterial, for either this or the next
mitosis must separate paternal and maternal elements. Figs. 3 and 4, representing respectively polur and lateral views of this division, show the eight
SO3IATIC CHARACTERS A N D CHROMATIN l{Ol)-LESGTHS.
11 1
tetrads aiid the odd chroiiiosoiiie, aiid iii the latter the odd chroliiosoine is seen
p:tssiiig eiitirc to oiie pole, while tlic ordinary
oliroiiiosoiiics are preparing for or actually
undcrgoiiig fiusioiiiin the equatorid plaiie.
Tho *ecoiitl i1i;iturtition division iiiiiiiediately
foilowb the first, i i n t l the coiiiples is coinposed
of eight or nine clirouiosoiiies, the difference
depending upon tlie odd chroinosoiiie, which
is fouiitl in ouly 50 per cent. OE these cells.
As in the c:ise of the speriiintogoiiial iiietapliascJs, each ordinary chrornosome is composed
of two rqual rods, :and the wniiie siee-relntionFig. ,j.--Second lulltl,rntion
ships :ire again ;ipparcwt. Fig. 5 is an example
mi tosjs.
of this nietaphase stmi froiii the polar aspect.
!rhe tr;insforiii:itioii froin sperni:itids to unripe and ripe spcnnatozoa is
siiiiilar to tli;rt, alrsatly described f o r these orgaiiisnis by niyself and other
writers. Tlic cliroiiinsonies becoiiic dissociated into iiiiiiutc gfiiuules, which
at first stain oiily sliglitly with the iron h~iiiatosylin; tlic :ippe:iraiice of the
'' centrosoiiie " is accompanied by clongat,ioii of nuclcus and cytoplasm, the
latter cvciitunlly constituting a long tliread-like tail.
DIBIENSIONS
O F THE CHROMOSOMES.
The di:iiiieter of :ill coiiipo~ieiit rods of tlic ordinary sperniatogonial
chroiiiosoiiies is -83 p, :ind thew consequcntly differ from oiie mother oiily in
length. The coniplcs is divirible into two groups represented respectively
by tlirec loug and five short pairs : this grouping accords with that of Davis,
and with my own u1)on otlicr iiienibers of' tlic genus. Moreover, thc Icngths
of tlic coiiilwiiciit rods of tlie five short pairs are respcct.ively 1.7, 2.1, 2.5,
2-9, and 3.3 p, :ind tlwrefore constitiite consacutive niciiibcrs o E :I series in
:irithiiieticul progression ; those of the t,lirce loiig pairs also belong to this
series, but are ;&ernate instead of coiiswutive, being rcspectivelg 5.0, 5.8,
:u1d 6.7 p.
Thc! tetr:ids or priiiiary speriiiatocytc! cliroiiiosonies caniiot be iiieasurcd
uccurately for tlieir outlines are irregular, but a careful study of the filaments
condelising during tlie preceding prolhase leaves little doubt tlint, they are
int1ividu;illy coiiiposed of rods of the above diuicnsions.
r 7
1lie secoiidary sperniatocyte coiiiples is tlle most favourable for the
riieasurwiient of choiiiosoiiics, since overlapping does not occur and in,dividuals ;trc coiii1)osed n€ iuiirs of rods :is in the spcrniat,ogoiiid iiiitoscs.
The di:unetcr of' the component rods of ordinary cliro~nosoniesis again -83 p,
;i~idtlie leiigtlis ;ire respectively 1.7, 21, 2.5, 2.9, 3.3, 5.0, 5.8! and 6.7 p,
i. e., identical with those of the 8l)ermatogonia. The odd chroinoao~rie,found
112
MR.
C.
F. c'. MEEK ON THE CORRELATION OF
in 50 per cent. of these cells, is again easily recognizable on account of its
great l)readt,li. I have -already produced evidence to show that rods composing ordin:iry chromosoiiies of organisins above and including Nemathelminthia have a constant diameter, viz. -83 g, and that their lengtlis
constitute members of a general series in arithmetical progressioi: ; and the
chromosome measurements of S. c i r d p e m i i s therefore afford further s u p p r t
to this ;mumption.
The accoinpanyiiig figure (fig. n) shows tlie complexes of Stenobot1wzc.r
pai*allelirz, S. vii*iCEzclus, 8. h i c o h , and S. owtipennis, the four complexes
being respectively marlred A, B, C!, and D. I n my recent paper I have
ident'ified rod-lengths of the general series by numerals, which are again used
and tire placed above the corresponding chromosomes. The drawings show
coinponent rods, e:ich sperrn:itogonial aiid secondary spermatocyte cliromosonie
being composed of two aiid eticli priinary spermatocyte chroniosonie of four.
The rod-lengths of the fire short chromosomes appear to be tlie same in all
four species, whereas tliose of tlie three long chromosomes are not identical
i n miy two : in S. parallelus they correspond respectively with Ohromosoines 14, 15, and 1 7 of the genernl series, in S. rii*idzilii.s with Chromosomes 13, 15, and 17 ; in S. / ) i c o l o ~with (Ihroniosomes 11, 13, and 17, and in
S. ntrtipennis with Chromosomes 11, 13, and 15.
If now we iissiiine t'liat the chromatin is directly conccruecl with tlie
transmission of thc hereditary characters-and
we ]lave many reasons for
assuming this-\w inlist look for tlie cause of soinatic differences bet\reen
thwe species in the Dhree long clironiosomes, for the respective nuclei
appear to differ only in the lengths of these. Moreover, we must try
to discover horn tlic$e differences i n rod-lengths have occurred, for the
problem of cliromossme function must lie intinlately connected with such
differences.
Let us firstly assume t1i:it cl:ro~nosorne rods tlirougliout the aiiiinal
kingdom are of fised lengths, and that niorpliologicnl siniilaritp is inrari:tbly
accompanied by functional similarity. This assumption carries with it tlie
furt.her assuniption thnt in the course of evolution certain rods hive disappeared from each coniples, 1oc:il c o n d i t h s having determined which
should persist and which should be eliiniiiated : moreover, it postulates a
greater nuinber of clironiosomes in primitive than in 1:iglilv orgitliized types,
:d
we must espect to find allied organisins possessing ninny chromosonie
lengtlis in coninion. The foriner of the last-namcd corollaries is, however,
not supported by actual investigations, and with regard t o the latter I have
already shown t1t:it E b ~ ; ~ c udoes
/ a not, possess one rod-length in coininon with
Steiiol~otlii~us. TVe are accordingly faced by a coinplete con tr:tcliction, for, if
:I definite cliromosonie rod-length is invari;iblp correlated with :i definite set
of somatic characters, no such set of characters can be possessed by botl:
earwings and gr:isshop~~ers-itiembers of sister fnndies.
SOMATlC UHAIiACTEItS A N D CHROMATIN ROD-LENGTHS.
'.
'.
.
113
114
MR. C. F. U. MEEK ON THE CORRELATION OF
The second and alternative hypothesis postulates a continuous linear
growth of chroinosoiiies in the course of phylogeny, and is based upon
data that support direct correspondence between the degree of soinatic
coinplexity of an organisiii and the total volume of its germ-cell chroinatiii ;
it inoreover off'ers a logical explanation of the evolution of various rodlengths. The measureiiieiits that I have given can be only approximations,
:ind the difference between terins of the general series is probably snialler
than that shown, but these ineasureiiients suffice to prove digerences between
coinplexes of allied species ; and it is only reasonable to suppobe that such
differences are of comparatively recent origin aiid have evolved by soilit)
coiitinuous process accompanying the soiiiatic differentiation of the species.
I n certain cases the process niay have been coliiIJlicated by degeneration,
possibly resulting in the couiplete disappearance of a particular clironiosoiiie
froin the coniples, but, even if this additional factor is eventually established,
I :in1 aware of no reasons for discarding the assuinptioii that the guiding
principle in complex formation is and has been a purely 1ine:tr growth of
coiiiponeiit rods.
1 have already pointed out that if this second hypothesis is subsequently
pi-oved, morphological identity of clii*oinosoniescan be no guide to functional
identity outside the narrowest limits of our classification ; but in the case
of allied species, which must have evolved tilong : h o s t piirdlel lines, we
may reasonably hope to establish correlation of iiidividual ~Iironiosonies
of the respective complexes, and thus forin a basis upon wliicli to :tttcnipt
correlation of rod lengths aiid definite somatic ~ I i a r a c t e r ~ ,
Let us therefore consider again the complexes of tliese four species of
Stenobotlirus. Tho lengths of the five short chroniosoines :ire the saiiie in all
cases, ant1 their identities appear consequently to be established ; the three
long chroiiiobonies, however, are not the same in any two couipleses, and
correspondence is therefore not at oiicc apparent. If' we accept the first
hypothesis, wliicli postulates inviiriable correspondence botwcon definite
rod-lengths ;uid definite sets of' sonutic cliiiracters, we must realize that uo
long chroniosonie is coilinion to all four species ; Chkomosoine 11 is absent in
6. parallelus aiid S. &*icluhs, Chroniosonic 13 is absent in 5: pavallelus,
Clirouiosoiiie 14 is absent in S. uis-idulus, S. hicolor, and S. eurtipennis,
Olironiosonie 15 is absent in S. bicoloi., and Ohroniosoiiie 17 is absent in
S. curtipemis, If, on the other hand, we accept the second hypothesis,
which postulates continuous 1ine:tr growth of rods, we must realize that in
these coniplexes the short, inediuui, iind ' long chroniosonies of the three are
probably respectively ideiiticul. Tlius Chroiiiosonie 13 of S. oivitkilus docs
iiot correspond witli Chromosome 13 of S. bicolov aiid S. czivtipemis, but
corresponds with Chromosome 11 of the two latter aiid with (:hroniosoiiie 14
of 8.yumllelus, for these cliroiiiosonies constitute the shortest ineiiiber of
the long group in each case ; siiiiihrly, Chroiiiosonie 17 of S. p u d l e l u s ,
GOhIATIC CIlAIiACTERS AND CHROMATI?; XiOD-LENGTHS.
115
S. i*Ivicitrlrrs, and AS.Licoloi. is not unrepresented in S. o r ~ t i p ~ i z i z i sbut
,
i4
functionally identical with C%roinosomc~ 15, which has not yet grown
sufficiently in tlie last-nanietl to be c l a w 4 in tlio higher category, C'orrespondencc between tlic reniaining long c4iroinosomes can be seen in thcb
diagr:iiii I)? following tlie :irrows, wliicli l i n r e been inserted to show identitie4
in all caws.
If the difference 1)etwecn ternis of the general series i4 eventually foiind
to he -mnller than half tlie rod tlinnirter, cliroinosonies that wc now class a s
c ~ i i ~ r ~ u tmust
i r e be separated by interinediate lengths, and tliose that we
c h i s together niay, :I* a result of morr accurate means of nieasiireinent, be
shown to ha\ c niinutely cliffwing lengtlis : tliic, Iio~vever,cannot affect the
three long cliroinosonies of ,kenoboil~,./rs,for their respective lengths are
such that nii~tnltei n identification is impossible. We inubt nevertheles.;
rciiienibrr thxt the corresl)ontlcnce indicated Ily the :~rrowsia based upon
a 1)iirc liypothcsi~,and tlixt :it p r w w t we possess 110 direct evidence in
sui~porlof this corrcspontlence.
Turning now t o the consitleration of bornatic characters, we find tliat t l i e ~ c ~
grasshoppers arc inclividunlly distinguishetl by slight but clearly defined
differences. Tlie following descriptions give tlie principal cliaracteristics
and arc quoted from tlic n.orIi4 of BoliIar, Burr, and Kirby.
STESOBOTHHUS
I'AIL~LI,EI,TS.
1. Vrry varinhle in colour.
2. Antenlip l o n p r i n niale than in female.
?,. Pronotnm with transrersc furrow nearer to postcrior tlinn to anterior
border.
4. Lateral carinz of pronotuni almost pzrallel or slightly approximating
about the niicltlle.
5 . Elytrn do not reach ~ i t of
l abdonien in male ; in female do not rcacli
1)qoncl fourth a1domin:il segment.
6. Medi:istiiial a r w of el+a estcndcd al)rnptlp toward,.; the apex Rut1
extentled round base, forming n rounded lobule ; anterior margin of
elytra conwx round Ime.
$. Wings rudiinent:wy.
8. J,ength of hotly, inale 14-15 mni,, female 18-21 mm.
~TESOBOTHIIl'S V I I ~ I D U L I ~ S .
1. Grecn varied with dnrltor.
2. A ~ i t e n nlonger
~
in male than in female.
:;. Pronotuni with transverse furrow midway I d w e e n anterior and postcrior
borders.
4. Lateral cnrinw of proiiotuni sliglitly angled near anterior border,
rounded slightly posteriorly.
116
MR. C. F. U. MEEK ON THE CORRELATIOK OF
5. Elytra fully developed in both sexes.
6. Mediastinal area of elytra gradually extended towards the apex,
prolonged to length of anterior border and not lobulate at base.
Anterior border straight.
7. Wings developed.
8. Length of body, male 13-15 mm., female 20-24 mm.
STENOBOTIIRUY
BICOLOR.
I . Very variable in colour.
2. AntennE of equal length in both sexes.
3. Pronotum with transverse furrow nearer to anterior than to postsrior
horder.
4. Lateral carin= of pronoturn sharply angled in anterior part, diverging
towiirds anterior and posterior borders.
5. Elytra fully developed in both sexes.
6. Mediastinal area of elytra extended abruptly towards the apex and
cstended round base, forming a rounded lobule ; anterior margin of
elytra convex round hase.
7. Wings drvcloped.
8. Length of body, male 15-16 mm., feinale 19-24 mm.
'STEKOHOTHRUR
CURTIPENNIB.
1. Variable in colour.
2. Antenna longer in male than i n female.
3. Pronotuni with transverse furrow nearly in middle.
4. Lateral carins of pronotum straight and parallel.
5. Elytra very short.
6. Mediastinal area of elytra extended abruptly towards the apex and
extended round base, forming a rounded lobule ; anterior margin of
elytrn convex round baso.
7. Wings very short in female ; in male equal in length to body.
8. Length of body, male 17.5 mm., female 17.5 mm.
Let us now try to correlate these characters and the chromosome rodlengths of the respective complexes. We will firstly consider the lateral
carinae of the pronoturn; these are parallel or slightly approximating in
S. parallelus, S. viritlulus, and S. cvrtipennis, but sharply angled in S. bicdor.
This characteristic is distinctive, for Burr has pointed out that by it alone we
can distinguish the last-named from the other three species. Now if correlation is evident, we must expect to 6nd corresponding chromosomes of the
same length in S. parallelus, S. viriduluz, and S. curtipemis, but of a different
length in S. bicolor ; if, however, we follow the arrows in the diagram we see
that no long chromosome fulfils these conditions, and correlation is therefore
not established.
sonixm
CHARACTERS AND CHROMATIN ROD-LESGTHS.
117
Taking :is a second example the niediastinnl area of the elytra, we must
expect to find corresponding cliroiiiosoiiies of thc sanie length in S. 1)urulIPZUS,S. I)icolor, and S. c u d p e t i n i s atid of a different length in S. airitluliis,
for in the three first named the area extend3 abruptly towards tlie apex
whereas in tho last it extends gradually. W c again fail to observe such a
chroino-ome ; and tlie saiiio absence of correlation is noticeable with respect
to the otlier characters on the list with tlie exception of colour, which is
variable and untrustworthy.
Morc.over, we are not niorc successful if we assuiiie the first hypothesis,
whicli i,ostulatrs ulicltiil1gi1lg rod-lengths. Disregarding the arrows in tho
diagram and considering correspondence to depend entirely upon length, me
find tliat the length of a n t e m % :tnd iinglc of the lateral carinze miiy be
correlated w i h Clirotnosonie 15 in S. puralleltis, S. riridulus, and S. curtipenitis ; but tlie fact that 110 other characteristics appear to correspoiitl
with chromosome 1ength.j makes justification for this assumption doubtful.
Furtliertnore, tho genus Stetio1~otJwtishas been divided by Bolivar and other
systematists into suhgenent, iintl these species are now classified ;is follows :t 'Iiortliipptis pa~ulleltis, Cliortlii1)1)tis c i w t ; p e t d s , Ornocestus riridrilus, and
StattrotEems /,icolor : it is noteworthy that tlie two whose complexes sliow
the greatest differences in rod-lengths should thus be classed together.
Four explanations may be put forward to account for this failure. Firstly,
we niay u~sunictliat my iiie;isureiii~iitsare inaccurate ; this, however, seems
unlikely, for great care has hecii eseroised, and thedlengths of the long
chroitiosonw~:Ire such tlint relative error should be impossible. Secondly,
we may ;ibsuine t l u t the lengths of the five short chromosomes are
not re*pectively identical in all the species; i u this case the characters
mentioned may be corrolated with tliese and not with the three long
c~1ir01iiosomes
: if, liowevw, the principtil somatic differences, upon which
systmiatists have based tlieir classification, are not trmeahle to obvious
differences in long ~ h r ~ i i i ~ why
~ ~ mshould
e ~ , they be traceable to imperceptible differences in short chromosomes, and, if they are so traceable, to
what are the obvious differences in the former due P Thirdly, the arrows
may be misleading : Chromosome 17 of S. p r a l l e l u s , S. viridiiliis, ,and
S. hicolor may, for example, correspond in S. curtipemis with Chromosome 13
and not 15, in which case the last mined chroinosoine corresponds with the
medium instead of the longest member of the three, and is accordingly
functionally identical with Cliroiiiosoiiie 15 of S. ~~urullelus
and S. &idulus
and Chromosome 13 of S. I~icolor. This is undoubtedly possible, if rods are
continuously increasing in length, for a long chroinosome may in the course
of evolution be overtaken and passed by one that was shorter, and the latter
may consequently be mistaken for Blie former : if this occnrs, we inust realize
that measurements cannot always be a trustworthy index to functional
correspondence even in the most closely allied organisms. Lastly, if the
118
MR. C. F. 17. MERIC ON THE COHRHLATION OF
chromo~omesof a complex are qualitatively different, as we have reasons for
believing, each muat be concerned with a definite set of ch:iracters : difference
in length of two corresponding chromosoines niay therefore be connected
with differences in several characters, antl, even if the character under
consideration is included in these, other factors are equally bound up in
the chromatin rod, aiid map be responsible for apparently irreconci1:tble
lengths.
I n the circumstances I a m inclined to think that the fourth explanation,
possibly coupled with the third, will eventually be found to account for our
present failurc. I t is difficult to believe that the obvious somatic differences
mentioned in our list are not in some way connected with tl10 lengths of' the
three long chromosomes, but until a thorough an:dysis has been ninde of both
internal and external chariicteristics of these species we cannot hope t o
correlate somatic characters rind chromosome rod-lengths in the genus.
Tn 1908 McClung indicated a course of investigation upon Acridiidrc to
he carried out upon these lines, but I have seen 110 paper by hiin on the
subject, : possibly he and his followers have been able to throw some light
upon this difficult problem.
RESLTMR.
Each ordinary sperlnatogonial and secondary spermatocyte chromosome
of S. cirrtipennis is composed of two equal rods, antl each primary bpermotocytc chromosome of' four. The diameter of these rods is invariably
-83 p and consequently lends further support to the assiimptioii that tlic
chromatin thread-width is constant iii all organisms ahove and including
Nemathelniintliia.
The lengths of the ordinary rods constitute ineinbers of a general sericlc
in arithmetical progression ; the five short chromosonies are respectively
identical with those of other members of the genus, but the lengths of the
three long chromosomes once more enable the species to be identified.
A coinparison between S. puwdlelirs, S. riritklus, S. bicolor, and S.czitstipennis fails to establish corrcliition of' somatic characters and eliroinosoiiic
rod-lengths, but we have reason for believing that the obvious characteristics
upon which identification is 1)ased are in some way connected with the three
long chromosomes : our present f:iilure is probably duo to ignorance of the
less obvious somatic characteristics and ti the lack of trustworthy methods
of identifying corresponding chromosomcs in the respective complexes.
SOJIA'I'TC CHAR.4C'rERS A N D CHROMATIN
119
ROD-LENGTHS.
BIBLIOGRAPHY.
C!/toloyical.
DAVIS,HERBERT
SPENCER.
1908. Spermatogenesis in Acridid% and Locustids. Bull. Mus. Comp.
Zool. Harvard, vol. liii. pp. 57-158, tt. 1-9.
GEBARD,POL.
1909. Sperinato,nen&se chez Stenobothrzis biguttulus. Arch. de Biol.
vol. xxiv. pp. 543-625, tt. 19-21.
MCCLUNG,C. E.
1'300. The Spermntocyte Divisions of the Acrididae. Kansas Univ. Sci.
Qu:irt. vol. ix. pp. 73-101, tt. 15-17.
1902. The Sperntatocyte Divisions of the Locustids. Kansas Univ.
Sci. Bull. vol. i. pp. 185-231, tt. 7-10.
190.5. The Chromosome Coinplex of Orthopteran Spermatocytes. Biol.
Bull. Woods Holl, vol. ix. pp. 304-340.
1908. Kylologp and Taxonomy. Kansas Univ. Sci. Bull. vol. iv.
pp. 190-215,253-262.
I I E E K , CH.ARI,ES F l i A S C I S ~JLLATHORNE.
1:)11. Tile Rperinatogenesis of Stenohothriis viridulus. Journ. Linn.
Soc., Zool. vol. sxxii. pp. 1-22, tt. 1-3.
2912. A Metrical Analysis of Chromosome Complexes, showing Correlation of Evolutionary Development nnd Chromatin Threadwidth throughout the Animal Kingdom. Phil. Trans. Roy.
SOC.SM. B, V O ~ .203, 1912, pp. 1-74, tt. 1-5.
SINSTY,12. 1 ) ~ .
1901. Ilecherches sur la Biologie et 1'Anatoniie des Phasmes. La
('ellule, xis. pp. 116-278, tt. 1-5.
i~~orpliological,
BOLIVAR, IONACIO.
1897. Catdogo sinbptico tle 10s Ortdpteros de la Fauna Ibbrica.
Ann. Sci. Nat. Porto, iv. pp. 105-135, 203-232 ; v. pp. 1-48.
B U I ~ RMALCOLM.
,
1897. British C)rthoptjera. Published by the Econoiiiic and Educational
Museutn, Hutldersfield.
KIRRY , WILLIAM
Foiis E LL.
1910. 9 Synoriyniic Catalogue of Orthoptern. Vol. iii. pilrt 11. (Locuutidrx? vel AcridiidE). Published by the British Museum.
LINN. JOURN.-ZOOLOGY,
VOL. X H X l I .
11