PRESLIA 1962

34: 127- 139

Askell Love

The biosystematic species concept (Institute Botanique de l'Universite de Montreal, Montreal, Canada)

Es ist mit Meinungen, die man wagt, wie mit, Steiuen, die man im Bretto voranbe­wegt: Sie konnen geschlagen werden, aber sie haben e in Spiel eingeleitct, das ge-wonncn wird. G

0 c t h

0

Taxonomy in all its phases is among the most fascinating of biological sciences. It is, also, the most important of botanical discip}jnes because its conclusions are fundamental to other approaches in general and to the study of evolution in particular. The impact of taxonomical studies in plants and animals on the theory of evolution is incalculable , since this idea was very strongly documented by observations of variation of living beings (LAMARCK 1809; DARWIN 1859, 1868). Antithetically, however, the evolutionary argument has of late become directed toward the definition of the taxonomical cate­gories. Although this first resulted in morphological and chorological defini­tions of the basic concept of species that produced an immense confusion as to the real nature of this category, the evolutionary argument has lately become the strongest supporter of a sound and scientific species concept: no other means of study is as effective in explaining the causal factors in the creation of the hierarchy of natural categories as is the analytical and synthetic approach to the study of evolution of all groups of biota.

When LINNAEUS (1751) decided upon the standard of species to be followed in biological taxonomy, he accepted a category which is a greater reality in nature than are other taxonomic groups and which is, therefore, most distinct to human observation. In selecting this standard, LINNAEUS followed a j udgment originating in dim antiquity. It was evidently his intention that the species of animals and plants ought to be of the same indisputable distinc­tion as are man and ape, cat, dog, horse and sheep, or apple and pear, barley, rye and wheat. Although he did not define his standard from this point of view, probably because he was convinced that the number of species must be limited since they were an act of creation, it is evident from his publications, especially on the Swedish flora and fauna, that he regarded the category of species to coincide largely with the cessation of hybridization possibilities or miscibility.

Because LINNAEUS did not feel a need to define his species standard very sharply, other botanists soon ventured to do this in order to obtain a distinct guide to determine this category. The definition most closely related to the wNks of LINNAEUS himself was phrased by DE CANDOLLE (1813), who re­garded the species as " la collection de tous les individus qui se ressemblent plus entr'eux qu'ils ne ressemblant a d'autres; qui peuvent, par une fecondation reciproque, produire des individus fertiles; et qui se reproduisent par la ge-

127

neration, de telle sorte qu'on peut par analogie les supposer tous sortis origi­nairement d'un seul individu." If this definition had been strictly adhered to by all biologists studying the variation of species even within the heavily explored regions of Europe, then no confusion in the concept of this category would have ever arisen.

It was the Linnaean species that was the basis of the theory of evolution as explained by DARWIN (1859), though he was also the first evolutionary biologist to become confused anrr deviate from the Linnaean standard. There­fore, he proposed a vague morphological definition of this category: "I look at the term species, as one arbitrarily given for the sake of convenience to a set of individuals closely resembling each other, and that it does not es­sentially differ from the term variety, which is given to less distinct and more fluctuating forms." Elsewhere in the same book, DARWIN even wont so far as to affirm that "in determining whether a form should be ranked as a species of a variety, the opinion of naturalists having sound judgment and wide experience seems the only guide to follow." This may appear reasonable to those knowing and recognizing the Linnaean standard, whereas it must be utterly confusing to those who try to follow the Darwinian definition of this category.

Botanists studying the variation within the Linnaean concept of species soon discovered that morphological characters alone are not always an easy means by which to distinguish species from the category immediately above or below. It also became evident that the limit between species is not always sharp and that, furthermore, some species include considerable morphological variations that may also be geographically or ecologically distinct. Even LINNAEUS (1751) knew this and included such variations in his concept of variety. When EHRHART (1788) realized that such variations within a species actually may be at different levels, he proposed the category of subspecies as a name for a major geographical race that is composed of varieties or minor geographical races. Although this was a very reasonable solution that did not require any change in the Linnaean standard, some students of these varia­tions preferred to deviate from the classical approach and use the species cate­gory not only for the Linnaean species but also for these geographical varia­tions. This trend had started somewhat earlier, but it reached one of its ex­tremes by the time that SENDTNER (1854) accepted as distinct species two morphologically very similar ecological races of Pinus montana only because he felt it unlikely that they could belong to the same species since each grew in such markedly different habitat, P. Pumilio in mires and P. M ughus on limestone respectively. The same tendency to use the species category for small splits from the Linnaean species, though retaining it also in the classical sense for other taxa, was strongly advocated by KERNER (1866) , who, quite correctly, pointed out that the difficulty to distinguish morphologically between the Linnaean species and its lower categories must be a result of a continuous evolutionary process. Instead of accepting the proposal by EHRHART (1788), however , he preferred to give each of such variations a species name, provided that they could be distinguished morphologically, described, and then again recognized. WETTSTEIN (1895, 1896a, b , 1898), who broadened the studies of KERNER (I.e.) and provided additional details concerning the chorological effects on the evolution of variations, continued the use of the species category not only for the Linnaean standard but also for closely allied, smaller taxa of

128

undoubtedly more recent ongm though chorologically somewhat distinct. Although the choro1ogical approach has strongly jufluen ced even moderate

taxonomjsts outside Central Europe , even ju Scandinavia, Britain , and the N etherlands where botanist s adhered most conservatively to tho Linnaean standard, it has nowhere reached such an extreme as in the so-called " KOMAROV School" in the Soviet Union (cf. J uzEPCZUK, 1958). The value of the ma jor product of the ideas of this School, the otherwise exemplary Flora SSSH, , is greatly dimiJlished by the fact that , in some genera, even small morphological variations showing some slight degree of chorological distinction, are classi­fied as species . The most starting example to be cited in this instance is the genus Astragalus with 849 so-called species filling a volnme of its own.

The morphological-chorological method was not accepted by an taxono­mists, alt hough the concepts of many have been confused by it. One of the strongest dissenters to these ideas was UELAKOVSKY (1873) , who firmly advo­cated both the n eed for a synthetic approach and the n ecessity of maintaining the Linnaean standard. This may be the reason why Czech botanist s n ever went to t he same extremes as did some of their Austrian coll.eagucs with the result that the flora of the western Carpathians has never been split up in to as m any microspecios as has the equally variable flora of the Alps.

It ought to be emphasized that the morphological-chorological approach is wholly based on the theory of evolution. J t s imperfection seenrn , however , to be cam;ed by overemphasis on comparative morphology corn bined with studies on distribution and too weak an emphasii:; on differential analysis com­bined with expArimcntal synthesis . This ii:; just what should have been expected a t a time when the theory of evolu t ion had not yet been studied from the point of view of the mechanisms and processes of evolution that wore later to be explained ·with the aid of cyto logy and genetics. Hefore these processes were di scon'rcd , rnorphoJogicaJ com pari f:·o n was the only method available. As long as it "\Vas kept separate from the obRenations on the ccasjng of hybri­dization possibi lities, so important from t he point of view of LTNNAEUS and DE CANDOLLE, a confused rn;e of the caLcgory of species could scarcely be avoided .

Although the KERNEn- W ETTSTEm- KOMAROV species concept has had greater influence on taxonomists t han any other con cept, f:avc t he Linnaean , many related definitions have been proposed in the past to lessen t he apparent gap between these two approaches. Those cannot be discussed here, but re­ferences to 1nany of these will be easily fonncl in t he fairly comprehen sive reviews by SEM1mov-ri1rAN-SHANSKY (IUIO) , D u l{rnrrz (ID:iO) , DoBZHANSKY (JD37, l!J51) , MAYR (1D42, 1957 ), MA NS FRLD (In4-fl) , CAMP & GTLLY (1043) , HoTHMALRH (In55), VAN STEENIS (1%7) , J uzErczuK (1058) , LAMPH1WHT (I %0) , BEAUDRY ( Ul60) , and GRAN'l' ( 1D60). Although differently enunciated and with e mphasis on various characterjstics of the species, most of these definitions first fai] to recognil';e the simple and basic biological characteristics of the Linnacan species standard and then , try to alter the standard to fit the definition , without reducing the confusion the least.

A more definite return to the Linnacan standard has been strongly accen­tuated by the more recent approach we name biosyst ematics . This is the study of the taxonomic categories and their distribution based on the analytic and synthetic methods of cytogenetics combined with the classical morphological and chorological endeavour. This approach has led to interpretations of the

129

species based on studies of the interrelationships and total pattern of each system of detectable components and the history of evolutionary divergence and its canses. It is from this approach that we have been able to conclude that there are four essential and independent processes of evolution: mutation , genetic recombination, natural selection, and reproductive isolation. The im­portance of each and all of these processes for the development of the characte­ristics of different categoricr-.; is becoming increasingly more ev idcnt. ] t ha8 also been possible to demonstrate that although an these processes arc of great significance in the development of the Linnaean species, reproductive isolation is mm;tirnportant because the emerging of the species a:-; such coincides largely ·with the occurrence of a barrier to miscibility (cf. 1VIAYR, 1042; LovE, 1960a, d). This luts made it possible to find a biological norm for a clear definition of the species standard of LINNAEUS based on present evolutionary knowledge.

The idea of defining the species through means of its reproductive isolation is not new - it was expressed already by HAY (1686) , and it also constituted an integral part of the DE CANDOLLJ1J (1 8 1:3) definition. R ecently, however, several attempts haYe been made to phrase it in such a way that it exprc8ses the results of biosyst ematic knowledge whi]e at the same time being practical. Ono such definition was proposed by TuRESSON ( 1D22a, b) for what he termed ccosp ccies, whose members are required to be able to interchange their genes without detriment to the offspring (cf. CLAUSEN, KECK & HrnsEY, 1940). Another such definition was coined by DANSER (192Ua, b , HlGO) for what he termed the commii-:;cuum, which is the total number of individuals that are connected genetically through miscibility (cf. VAN STEEN IS , 1!)57) . The clcar­rnit and most practical biosystematic definition , and yet at the same time equally applicable to animals and plants, was , however, formulated by 1VIAYR (l U40) in connection with his studies on hi.rds : " A species consists of a group of populations which replace each other geographically or ecologically an<l of which the ncighboring oneR integrade or interbreed wherever they arc jn contact or which are potentially capable of doing so (with one or more of po­pulations) in those cases where contact is prevented by geographical or eco­logical barriers" . A little later , :MA YR ( J D42) rc<lucecl this definition to the phrase : "Species are groups of actually or potcnfaally interbreeding natural populations, vvhich are rc prochlctivPly jsolated from other such groups." A better and more precise guide to the standard selected by LINNAEUS and hiH predecessors for the natural and distinct category of species has yet to be jn \·cnte<l. :Furthermore, a stricter adherence to such a definition would soon result in a distinct lessening of the confusion that has been caused by the lack of such recognition of the Linnaean standard.

It is evident that exact adherence to the biosystematic definition of the Linnaean species is possible only in thorn groups in which the occurrence of or lack of a barrier to reproduction has been established, or in taxa that arc available for biosystematic Rtudies. However, since biosystematic experiments have revealed that some character corn binations arc more indicative of the occurrence of a sterility barrier than are others, an experienced and critical taxonomist, when studying such species, is soon able to deduce fairly correctly what taxa are species rather than subspecies , and vice versa. 1~his is, in fact, a method fairly similar to that used by LINNAEUS and his followers. Even though they knew little about reproductive irnlation and its mechanisms, their skill in detecting real limits between species was so profound that bio-

130

::iystomatic approaches only rarely wi11 be able to improve much upon the con­clusiorrn they reached with the species they knew well. Whenever such 8exnaHy reproducing plants become available for biosystcrnatic stHdy, however, the conchrnions from the rnorphological-biosysternatic approach will be open to correction or, more likely, confirmation.

There are some groups of plants that never will become available for studie8 from the lJjoRystematic point of view. These arc the taxa that arc characterized by variom; forms of apornictic reproduction. To classify them on the bards of re productive iwlation would lead to a confr1Hion eYen greater than that created by the morphological-chorological method of study of these gr011ps, since every individual is reproductively isolated from all itR relatives. However, a taxonomiHt supporting the biosystematic species concept ought to be able to Rclect in theRe groups as well combinations of morphological characters that differentiate the real species from its lovvcr unitR and form cliscontinnities comparable to the reproductive gap of arnphimictic groups. This possibility haH recently been ventured by LovE (1U60c) and LoVJi~ & Luv:FJ (1U6lb) for Rome such taxa from central and northwestern Europe. 1t is likely , however, that all such attempt8 will be defied by those few who prcfor to split such groups into a logion of microspecies founded upon morpholog jcal and choro­logjcal disfanctions on 1y.

Strict application of the biosyHternatic definition of the Linnaean species iR possible in many groups. It is applicable to most genera in the floras of centrnl and northwrste rn Emope (cf. LovE & LovE, 106lb) , Japan and eastern North America which have already been studied from these points of view to a consi­derable degree. Here studies of the development of reproductive i~mlation have shovm that this stage is reached in either of two ways, abruptly or gradually. Jn the former, the irrcvernible reproductive barrier is formed suddenly, itSually by a euploid change in chromosome nnmbcr, and later followed by further morphological and choroJ.ogical diffcrcntia,tion by aid of the processes of mutation, genetic recombinatfon , and natu ral st>lection. In tho latter, hmvove r, the establishment of the disconfanuity is a contirnt0us and slow pro­cesR that will be mirrored in gradual morpho]ogical differentiation that is 1 ilrnly to pass through the lower intra.specific categories bAfore reaching the ~ pPcics limit (VALENTINE, l!J4D). Well-established gradual species have deve­loped a reproductive harrier so strong that hybrid s arc either absent or difficult to obtain, whereas others may RtiJI be able to hybridize and exchange some genes, though usually not without some detrimental effect on the progeny. In still other cases, however , it may be djfficult to distinguish , even experiment­ally, between gradual species, which are partially interfertile, and subspecies, which aire partially intersterilo, since Loth may be equally di stinct morpho­logically. Abrupt HpecieR, however, are strongly irnJatcd reproductively, al­though polyploids at different leve] s may be crossable. Because abrupt species , like the catcgorie::; of the gradual species, may be of different ago and then likewise formed in different ways, they may show morphological distjnction of all degrees from those characterizing formao of the morphological species concept through those comparable to the characteristics of varieties, sub­species, species and oven genera of the group of gradual species. To classify them ~t these levels, as proposed by some botanists (cf. HEYWOOD , 1958, 1960; BocHER, 1060) , js, however , fa ll acious because of their essential character of strict reproductive isolation. Irrespective of the degree of morphological

131

and geographical distinction, a forma, variety, or subspecies that is reproduct­ively isolated from other such taxa of the same species is an absurdity that is incompatible with the principle of evolutionary taxonomy and which directly violates tho Linnaean standard of the category of species (cf. Du RrnTz, 1930; VAN 8TEENIS, 1D!57); VALENTINE & LOVE 1958).

The reason for abrupt species showing various degrees of morphological <liHtinction is in part connected with differences in differentiation that have followed the formation of the reproductive isolation. It is, however, also due to the different methods with which such a speciation can happen. In general, abrupt species are formed either by the direct reduplication of the same set of chromosomes, or by the addition of different ones. The influence of this on the morphological distinction of the n ew taxon is not the same, though some morphological differences are always met with even in experimental poly­ploids (cf. :MuNTZING, 1!)53, 1955; TISCHLER, 1051 , 1953- ID57).

It is known t lrnt t.ho first event, autoploidy, or rather its phase panautoploidy (LovJC & L ovE, 194-!)), whiC' h is ifw dnplication of t h e same sot o f chromoso m es in a more or less pure-bred p o pu­lat ion . is rare n.s un cffnd ivo sp rwies producer, a l t h o ugh pana utoploids arc proba hl y form ed nt t h0 r ate of a frw caHcs in a thousand in mos t p o pulationH (of. L ()v 1;;, l 944b, 1!)60a, d ; Bow DEN,

rn+H; KrnLLANrnrn. 1953) . Hut t.li ey rarl'ly s urvive in Hature for moro than a s in g le o r a few gen er­a tions, twcausr' t his kind of chrom oso m o doublin g is often followml hy profourul m eiotic disturb ­ances, so tlwy aro of litt le s ig nificance in evolution. an d a r c iaxonomi<~all y u n irnror tnut (U)vE, I \J60a, d ). To g iY C) th e m a (",n~o nornic name iR, therefo re, h a rdly a< lv iHalil n. Rin co t lwy w ill nuL c\·en fil l th e r0quirnment by K.1-; trnER (L 866) t l1 at Lltcy should IJe rncognizable later. This seem s, h O\\"O\·e t', to h a\"(\ l>C'c n d()nO by Biic HL<at (l !)!)..j.) whon naming the variety duple.'"C of I'ulsatilln pra ­tcnsi8 on the basis of it s d o\·iatin g c hromosome number though only occurring occasio nally in sor:nc oLhcn,·i:-;c nu1·111 a lly diploid populat ions . Them a r0, h owever, known cases ofpan autotr iplo id y in wliicli s uch i11d i\· iduals n'p rocluce veg<'Lat i\'e ly a nll are t hus a ble to survive a11d even sprnad. In the s1H' r ics !J11fom11s mnbellatu s such trip loids arc gen orally ig n ored by taxu n orni Rts though t ll<'.\ ' a rc fairly \\"iclPsprc::H I (f'f. LoHAJ\Tl\1A K, I !ml, l!) fi4). In Po7wln8 tremula, tlrny h ave bee n namod rn' \·ar. uirws (cf. 1-lYf,AND l<.: H, 1\)40), wliorcas in Acorns, the Linnacan sp cc> ies A . Cnlamvs is su 0h a triploid that lin.s di s pe rsf'd m ·<' I' wid e a r <'as hy a id of huma n agonci<'s (VV1JLf<'F, LH40, 1\)54; J ,i)vro; & Lii\-K J Dfi7a. b). Tlwso tl' ipl o id s arc, in fact., not taxa at the Linnacan s p oc'. iOR kvel a l­tho 11g l1 th ey mn,v havo gain ed ::;01110 rnorp l1 o log ical di s tinction and r cproclndiv!' isolaL ion t hroug h thei r un lmla1wcd cy tolog ical con diLion and also h ave form ed a g<'og rap hiea l arna of t heir own. 'l'h0y are rat lt PI' Lo be eornpa1·e c/ w iLh some aporn ieLs or, equally co rroetly, with hybr id s that uro perpctuat.cd tl1rn11p;h n lgotativo roproducLion. lf tlrny arn to b e g iven a taxonomic rank at a ll , this 011gl1t to IH' l' ithcr t hut or nn agamo\·arict_r o r , in t l1 0 case of Acorns a rnl ot.lior s n c h Lripluids wilh so 11H'W irnJ llll<'Prtain pHr<'ntagn, that of a h ybrid w ith a bin a ry name.

In ran• cas(';-;, t llf' ro is so m e possibility of rapidly reduci11g t.110 disad vant.agoH of p a n a u t.o ­p lo id oR IJy s11rne se lndin' o r r eco mbinatio n procossoH. Then tlrn now a utopl oid w ill cl ovolo p in th(' din,diurt of tho other event (<'f. .l\ I UNTZ I N<:, l!l4:1b) anrl lJroduco what w o name an lwrniautop loicl, wh i<·h 1nay su l'vi ve a nd rorrn an area o f its own. H crn ia 11 toplo id s arc, othor ­wis<', prodtu·('d aft<' r h yb ridization nf somewhat different races o f tho same sp of' ios w lii ch have already diffe n ' ntiatcd Lh e ir chromosomes e nough to rNluco t il e m oiot ie disturhan<..:os b eyo nd the crili0al level l0tlrnl to pann.utoploids.· He 1~1iaut<>p lo ids h ave IJeo n oxporirn cnta ll y produced in HC' \"(' rn l eu lti ntLed pla11Ls (cf. TLscHLEll, J 95 1, H1 53 ·- L!Hi7). It is li ke ly t h at IJactylis glomemta ( ;\{(j~TZ f N(;, l g;n, L !H3a: Ho RH.ILL, 1901) , I'hlcum vratense (No IW 8NSK Li)L I>, J H4fi, Hl53), tho polyp loid se ri es A r·rtoselln (Li)v1.;, 1B4-3) , Antho.:canthum odoratum (Owr1c 1tcrn.r:N, 1942; RozMus, .I 90!1: L iin; & L'''1v1;; , H15G) and i-;evcra l oth ur g ro ups we ro originall y fo rmed b y t his procoss. Tl1i s, liowC\'N, ran n ot a lwayB b e eas ily ve rified bocauso the li mit between h cm iautoploid s a nd ccrli1i1t p lan Ls belonging to t h e oth er procesH is n ever s harp.

The other event, or the addition of different whole sets of chromosomes, is named allop loidy. It has certainly been a much more important factor in evolufaon than has autoploi<ly (cf. STEBBINS, 1950). The chromosome sets of allopJoids may derive either from clearly different races of the same species in which differentiation of the chromosome sets has already commenced, or from different species which are properly isolated reproductively even

132

though they are characterized by the same chromosome number. When the parents are only racially distinct , then the original hybrid from which the taxon with the double chromosome number derives, will be only partiaJly sterile. This is the only characteristic that differentiat es them from hemiautoploids , which) by definition, derive from hybrids without even partial sterility. The original alloploids that have been formed in this \Yay may have some of the properties of autoploids, although all such negative properties will soon be selected away. They are named hemial1oploids, and they are supposed to be the most common kind of abrupt species (cf. STEBBlNS , HH50, lDGO). Many hemialloploids have been studied by biosyskmatists . From the taxonomical point of view, the legion of examples of su ch t,axa t hat could he mentioned vrnuld range from the unequivocal to the controversial: they exhibit all gradf's of morphological distinction and all sizes of geographical distribution areas, depending upon their age and the disfan ction of their p arentR. Although they sometimes may display only minor morphological differences, thev all certainly fill the basic requirement of the Linn aea.n species since they ·}iaye acquired not only morphological and geographical distinction but alrn a strong repro­ductive isolation that permanrntly preventR their intcrmiscibility wit h other such t axa.

When t he parents of a.lloploids are already distinct species which h ave differentiat ed · their chromosomes so far as to make p airing between them difficult or even impossible, then the first hybrid wm be very steril e and some­times very difficult to obtain , whereas the plant wjth its du11licakd chromo­some number will be perfect ly fortilP, or nearly so. In other instancrR the parent species may have reached such a degree of differentiation (other than poly­ploidy) that hybriJization is entire ly excluded , though this barrier may occasio­nally be surpassPd by a one-step proJuction of an a l loploid from t he vPry rare coin ciden ce of a ferfa lization of an unreduced female gamete of one parcut by an unredu ced male gamete of the other, sincP such nnreduc€'d gametPs some­times are able to break t hrough an incornpat,ibility barrier that is lm br-atab le by the reduced celh; (cf. B E JrnSTROM, 1H53). These proceEF:cs arc wl1at is termed pana11op1oidy, and it may wdl be of more frequent occurre nce t han we i::us pcc t at present. Jt is very cffecti\'e in the inFitani anco11 s cYolution of sections , or even of genera, and is widely recognized as a producer of distinct s11ecies.

J\Jo:;; t pnrrnlloploid s arc i axnnomic·al ly 1111 mnhig11ouH. · ns ::tr<' t l1 <· rceen tl y tlc-sc r ihed gl' nus lfylandrn ( Li)\" 1·: . l!JGl ), a nd U1 e ihrcc wL·ll -lrnow1 1 spec ies o f" 'l'ritic11m («f. N 1 ·~AHS, l!l48, JnGU; Liiv:c: & Liiv1~, l!Hi l a, b: NATO\A lt & N·n:uu1Ns . l!J lifi; ('1-11 ,:NNA \'1•: 1•: nA IAH,. L!lGO ), an d 8purlina 'l 'ou·nsendii (H l1 S1'TNS , l!J~{ I ), as wP ll as t he PX[Wrimcn t n ll y n· rifl Pd Oa /eo71s1:s T etrn hit (J\TijNT­

ZTl"G, Hl30, IU32), 1\'icot.ir11w 'l 'obnn 11 11 («f. UooDs n :ED, JUG4), rrn d JJrassirn. Nu7ms (FnANl>SKN,

L!)47). BuL som(' nm s till <·ou t ron'rs ial bcf'a11 Ht' th fly exhib it diffe rnneos so rni111 1Lo tli aL t li c, ir origirn1l dd ccLors lw d diff i<·ult,y in di s tingui sh iug Lbcm .. as, e .g., in L eymvs arenar·1·11.s wli ie l1 orig i-11atflS from f,, mo/lis and T.1. 8a u11los 11 s (ef. B ow1n:N, lflG7) , a nd in l\. oh1rU1(8ch ia Nan fe11ilii w hi ch apparenLly originatcR from Llio hy brid between J(. 7>rolifem an d J(. vel1.1tino (cf'. D ll HN AT, 18!)2; :H t<;YWoou , ] flGO), a lt hough it is u s 1H1ll~r r<'ported as rrn ly a polyploid "raco" of 1\. 71rol1fera (Bii ­c11En, L A B S l~N & RAHN, Hl fi3; LAw-;1.:N, HlGO) . To c lass ify o it lrnr t ho parents o r sueh po ly ploid8 as races only of Lb o a lloploid, or the a ll oplo ids as a race of ono of tho par0nts, ii;; noL on ly taxo­nomically un Ro tmd and against tho req uirements of ihe Linnacan s pcc i('R stand ard , hut a lso dirc,ctly a violatio n of common sonso a nd log ical taxonorn ieal principles , as so eo r rcc tl y d emon­st1·atcd by .M UNTZINC: (1 \.l30) in connoet ion w ith hi s di scw;sion of t h o morpl1ological c lassificat io n sch e mes p rop osed by so m e ea rli er taxonomis t s Rtudy i11 K Lho subgenus 'l'etrahit of Galcopsis.

In the group of abrupt species, the application of the biosystematic definition of the Linnaean species standard must inavoidably produce an increase in the number of species, since taxa differing in chromosome number but ignored or

133

classified lower because of morphological inconspicuity will have to be raised to specific rank (cf. NANNFELDT, 1938; LovE, 1951, 1D60d). In the group of gradual speciE~s, however, the effects of the strict application of the biosyste­matic definition will greatly reduce the number of taxa classified at the species level , since many post -Linnaean taxonomists t ended to be deceived by the conspicuous morphological and eco]ogica,l differences between some taxa that have not yet developed even the slightest degree of reproductive isolation. This is especially applicable to Central Europe where "there are still distin­g uished a considerable number of species which do not clc:::;crve that rank" (VAN STEmNLS, 1D57) . There are, however, also cases of gradual species which have been generally classified a,t a lower level than appropriate, as r ecently Te'J~.a,\_~cl t\\.\'O\.\~.).l. \)1-os~;;ten\.at1c extim·imonts, s o that oven within this group \i\le n1;>:-..XI ~;~\_(\~n_e,e \\Y.:.\J' \:liOmot1-mes Te-s-u\.t \n a;n incrcase(l rrnrnbor of species. 1\ O\.\g)lt 'C\\so to be i::mp\,aRi'Loc\ once more that the gradnaJ differentiation itself m\1st n ecessa;ri1y imply that Rome tax a will he fo1md t.o be at the jnterrnediate level so that even thorough experiments will be unable to conclude cleciRivoly if they are to be classified as species separated by a certain but incomplete degree of reproductive isolation , or if they are only subspecies which have gained a rather high degree of partial but variable sterility. Such cases have to be decided upon incliYidually, and some arbitrariness in their c]assificatjon cannot be avoided. This, hovm ver, does not matter much as long as the facts leading to this situation are not concealed or allowed to escape atteution.

A much discussed case that ma,y belong to the intermediate category just mentioned concerns the Linnaean species Geum rivale and G. 1J,rbamim and the taxa of the same complex later described (cf. GAJEWSKI, 19G7). Ecolo ­gically, those arc completely different. They also occupy somewhat different geographical regions . However, in places which ANDEI~SON (rn48) would have called " hybrid habitats" and al so under experimental conditions, hybrids arc easily formed and the differential characters of both ta,xa scorn to break down into a complete Aeries of interrnecliafos (WrnGE, 1926 ; MARSDEN - JONES, 1930). Although VAN STEENIS (1057) agrees with MARSDEN-JONES (l.c.) that they are classified too high at the species leve] , the cytogenetic evidence still seems to be inconclusive as to their real miscibility . Though it is also the feeling of th e present writer that they may be more correctly classified at the sub­specific level , it is hardly possible to claim it incorrect to retain them as Rpe­cies. Therefore, until further investigations yield more adequafo information, the continued u se of the arbitrary judgment of LINNAEUS on these Geum taxa can as well be recommended.

It is (1 different matter with tlrn two s pecies p n, irs Mclnndrinm r1lbr11m and "!'.f. album and Silrne vulyaris a nd S. mriritimrt . l<:x 1><'ri1rn·n ts 'vith the former p a ir , rarri ed out hy a number of inves ti ­gators , lifl,\'e clearly shown those taxa 1,o be <:ornpl e tely intorfortil<o nnd without even a vest ige 1)f

r0produ c Live isolation. l-l <' n c0, D. L6v 1-: (l!J44) condud<'d thattllf'y ar0 rorrcetly classifiod a s tho two su h sr oci0s rlioer:nm. a nd album of tho sper ir•s M. dior,cum.. This f'Otlf'l11sion is in oo nf'orrnity w ith the Linnaoan c lass ifi cat ion of both taxa as varioLiof-l only o f thn si nglo Rpoo ios Lyrhnis dioeca. Tlic iSilcnr, p a ir has hfl<-m m ost inte nse ly stucli ()( l hy MA1ts1J1<; N-.JoN1~s & T rr 1t1n1 •L (HJG7), n,nd although they Rt ill regard both taxa as difforont sp ec ieR , th e ev id ence as r oporLod by Lltcrn cloar ly support s i110 co nclusion by VAN NTE1<:N1s (1957) and Lc)vE & L<i vE (In6la, b) that they aro nothin g lrn t two subs pec ies of 8ilene vnlyaris.

In t ho gen us Jlretosa, exporirnonts carried out s ince 1939 mninl y by the present writer have sh own t hat th o fonr taxa A. prnten sis, A. alpP,sfris, A. nivalis and A. am.bi yu a sLill have not d eveloped a reproductive b a rrier, although the last two seem to he n earin g the critical level of partia l interstorility towards the other r elated taxa and A. alpcslris shows groat racial differen­tiation at the varietal level. Fro m the biosystematic point of view, the four taxa arc , thus,

134

cor rectly class ified as s ubsp ecies of the species A. 7Jra l c11s1:s, whereas t.lie related A. tliyrsij1ora ap paro 11 t ly h as pruducel l a mo l'O distinct r cprndn('tive banicr so it iR appropriately soparutrnl at Lho s p uC' io:-; low! (e f. Ui v 1~ . 1944a, 1!)4!) , 1904; Lh v 1 ~ & Lii v 1·: , I 048 , l!)()J n,, b; N\\ 11•:TL INS KA, HH30).

Jn1111m ora lil o ('X p o ri111 onts with t,ho rn any taxa gin'u sp uc iC's w111ws within tlio ett!Lini.tcd w h oats lia\·e rnvPnlcd that t.ho on ly rcpro<lu c ti\·o liarri crn 011l·u u1t! 0n'd lwLwt'l' ll t lw <lifkn·nt kind :-; of whr'ats arn tlte pan tdloploid diffe re rwos in clirn1nosomo 1111111l wr ln•hv<'Cll Llw tl11·no g ood s p oc ios Triticll'rn 111011ococr11m, 'I'. turqidmn, and T . ucstir11m in 1 lw ir wirkr sense . All Lli e otltnr taxa oug h t, thcroforn , to b e g roupod und e r Lhcso sp e(' ieH a s subsp cC' il'R 01· r acr•s uf' lower s ig ni ­

fi oan co us has recently b oo n a<lvocakd by i\I AC' KEY (l!J G+), :-lEAHS (l!JGH), Bo\H>J·:N (L!J G!J) , C1rnNNAv1.; 1 ·~ ttA I A II ( 1060), and Lh vF: & T.,«jyp, (HJGla, IJ).

A need for a s ir:uiln.r reduction of sp 0c ie;::; to tlm 10\·(· I of s nbspoei0s iR indi<'akd in t .lw rcc·ont, r ep o rt on tlw 11111ny t-axa of /Jiantlrns studied b y CAHOLLN (I !l!'i7). Tl w sa 111 0 i s al :-;o Llin cns<' wiLli n1any ot.ltc r w11wra, c:-;p ociall y in Contra ! ~uropo wlwl'f' s pliLting rn1 thn bas is uf t.lie morpholog i('al­c lio rolog ical a pprntwh !ms beon mos t ev ide nt , altlwuµ-li t'.\'. pf' ri11H'nLal inn 1:-; t igal io n sL ill is lae kin g for many of t l1 osc grnups.

Although most of the gradually developing taxa just mentioned arc ecolog­ically rather than geographically Jiffercnt , some taxa that arc geographically highl y isolated so that they must have been separated over a Jong period of time have a lso been found to lack ropro<l.uctive isolation. This ir:;, e.g., appa­rentl y the case with the pair of Linnaean s pecies Platanus orienlal1~s, from south east ern Europe and western Asja, and P. occidental is and related taxa from southeastcrn North America (WINGE, UH 7; SAx, 1H33 ; STEBBINS, rnr>O; PARRISH, 1957; PrzzoLONGO, 1058) , so they arc probably more correctly placed as subspecies of a single species. A similar observation has recently been made in Xanthium struniari1.tm by LovE & DANSE RJDAU (Ul5D) and LovE & NADBAU ( In61 ). This cmnplex species seems to be made up of two variab]c subspecies, originating from the Mediterranean an<l. Central and South America respecti­vely. When the varieties of those subspecies from cljfferent parts of the world were brought together as weeds in North America and Europe, they h ybridized frec]y and, by aid of pronounced endogamy, gaye rise to more or less distinct character combinations that have greatly confw;;cd taxonomists following the morph ological-choro logical approach .

Such examples of gradual1y developing taxa that have been incorrectly classified by mearn; of t he morphological m ethod are rnrna.J]y of the kind mentioned above. There arc, however , al t:o cases in which biosystematic evidence has showi1 that low deg ree of morphological differentiation , though combined to a high degree of reproductive isolation, has som etimes led morpho­logical taxonomists to classify good gradual species as varietes or subspecies on ly. Thi s is evidently the case with the subspecies palustre and constrictum of Galiitm palustre, whi Gh, accor<l.ing to CLAPHAM: (1040), have d eveloped a. st rong reproductive barrier. The latter is , therefore, more correctly c]assified· as the species G. debile which has, most likely, evolved gra<lually~from the same original stock as has G. palustre proper. Another such example may be Oryza saliva and its so-called race japonica, which arc distinctly inten;;terile ( cf. Hsrnu & OKA, 1958) , altho ugh the meiosis is qnito normal. It is, however , premature to draw the conclusion from the available ev idence that thc,:.;e tax:.t ought to he separated as distinct species, as recently pointed out by KrnAHA ( l H59) in connection with a review of the prei:;;ont knowledge of the origin of cultivated rice.

It is remarkable that the standard for the species category , which LINNAEUS

( 1751) accepted as a concept for the most important entity in the natural hierarchy of the living world, has not weakened with time and progress of discovery. It was invented by the original method that is inevitably descriptive and designed to answer the basic question about what arc the facts ( cf. LOVE,

135

I D60b ). It changed only a little when the descriptive approach .was supple­mented by the comparative, since the distinction of this category as compared with ist next lower and next higher groups also was evident from these points of view. When the implicit postulate of physical relationship that formed the basis of the comparative method became explicit with the acceptance of the fact of evolution, the result was a phylogenetic classification. Since students of the differences between the categories in this system regarded the species category only as one link in a long chaiu of gradually evolving groups that pass successively from one step to another, it was perhaps natural that they failed to realize that one of these steps could be more important than the others; this was also influenced by the observations by the morphological method that species show all degrees of differentiation. It was, however, the more recent and new procedure of differential analysis with experimental synthesis, which is carried out by biosystematists , that was able to reappraise the significance of different categories based on studies of the causes of their distinctions. This approach has restored the antique standard of species to its former strength as a group of indisputable distinction in the natural hierarchy. It has also demonstrated that although this stage can be reached in different ways, all good species of the Linnaean and biosystematic standard are char-. acterized by reproductive isolation from other such taxa.

Biosystematists are fully aware of the fact that al l material available to taxonomical studies has not reached the field of exact scientific evaluation that is typical of the advanced analytic and synthetic approach. They believe, however, that the classical standard of the category of species can and ought to be aimed at by those taxonomists also who need to follow the empirical art of classifying a single specimen representing the only knowledge of the living beings of a remote land. Such approaches may fail to observe the real distinct­ions of a species in some cases, but later studies will bring us closer to the truth. The only requirement biosystematists feel ought to be made is that whenever a more advanced stage of knowledge of an animal or a plant is reached, this ought to modify less exact classifications based on less advanced methods. 11 his is the simple principle of progressive taxonomy. Other procedures must be branded as reactionary and unscientific.

The species of LINNAEUS was that of the obscure past without an exact definition. It has long been evident that species show different degrees of morphological variation that may blur thefr real distinctions. The species of the biosystematist still is a variable unit that has evolved in diverse ways. It has been the privilege of the biosystematj sts not only to provide explanations for the processes of evolution of species and other taxonomical categories but also to discover that the only com'mon denominator for all good species of animals and plants, irrespective of their age or evolutionary history, is their reproductive isolation from other such taxa. The definition based on this knowledge characterizes not only man and ape, cat, dog, horse and sheep, or apple and pear, barley, rye and wheat , but also all other species that coin­cide with the standard that came to LINNAEUS out of dim antiquity. Jn that way , the n ew method not only adds evolutionary explanations to the concept of species but also maintains the continuation of this concept from times longer than man can remember. Thisis one of the reasons why taxonomy, the oldest and most important of biological sciences, remains among the funda­mental and most fascinating of all approaches to biological knowledge.

136

L ·i t e r a t u r e c i t e d

ANDEHSON E. (1948): Hybridization of the habitat. - Evolution 2 : 1- 9.

B EAUDHY J. R. (1960): The species concept: its evolution and present status. -- R evu e Canad. Biol. 19 : 219- 240.

B EHNSTROM P. ( 1953): Increased crossability in Lamium after chromosome d oubling . - H e re­ditas 39 : 241- 256.

Boc m o:R T. W. (1954): Natural populations of Pulsatilla in Zealand. - - Bot. Tidssk r . 51 : 33- 47. - (1960): Experimental and cytological studies on plant sp ec ies. V. The Carnpanula rotundifolia

complex. - Biol. Skr. Dan. Vid. Selsk. 11 , 4 : 1- 69. B t)CHER T. W., LARSEN K. & RAHN K. (195!3): Experimental and cytological stu<lios on plant

species. I. J(ohlranschia prolifera and Plantago Coronopus. -- H e rcclitas :rn : 289--304. B01m1LL M. (19Gl): The pattern of morphological variation in diploid aud tetra ploicl Dactylis. -

Jour. Linn. Soc., Bot. 56 : 441- 452. BowDEN W. M. (1949): Triploid mutants among diploid seedling populat ions of Asimina triloba. -

Bull. T orrey Club 76 : 1- 6. - (J957): Cytotaxonorny of section Psammelymus of the genus Elymus. - C[\,n a d . Jour. Bot.

35 : 951- 993. - ( 1959): The taxonomy and nomen clature of the whoat s, barleys, and ryes and their wild

re latives . - - Canad. Jour. Bot. 37 : 657- 684. B uitNAT E. (1802): ]lore d os Alpes Maritimos . Vol. I. - Conuvc & Balo.

CAM P ,V. H. & GIT.LY C. L. (1943): The origin and structure of species. - Bri tto ni11 4: 323- 385. CA ROLI N R. C. (1957): Cytological an.cl h y bridization studies in the genus Dianthus . - New

Phy Lol. 56: 81- 97.

CELAKOVSKY L. (1873): -Ubor d0n B egriff d e r Art in d or Natmgeschichtc, insh csond or e in dor Botanik. - Osterr. Bot. ZoiLschr . 23 : 232- 239, 27 1- 280 , 3la- 31 8.

C1mNNAVEEHAIAH M. S. (19GO): K a ryomorphologic and cy totaxonomic studi os in Aegilops. -Acta Horti Gotob. 23 : 85- 178.

CLAPHAM A. H. (194\1): Taxonomic prohlem s in Gali'um and .Juncvs . - In A. J . \V1LMOTT (ed.): British Flowerin g Plants and Modern System atic Methods : 72- 74.

CLAUSEN J., K ECK D . D. & Hrns 1~y W. M. (1940): E x p erimenta l s t.ucli es on tlio nature of Rpec ies . I. Effect of varied environmen t on west ern North Ameri<'an plants . - Carneg io Inst. l'ubl. 520: 1- 452 .

D ANSER B. H. (1929a ): Uber clio ni ede rHindisch-indisch en Stachytarpheta-Artt' n und ihro Bastarde , ue bs t B etrachtungen Li.ber uie B ogrenzung d or Arten irn A ll gcm cinon . - Ann . J a rd. Bot. Buitenzorg 40 : 1- 43.

- (l929b): Ub0r die B ogl'iffe Kornparium , Kommiskuum und K onv iv iurn uncl iibc r clic E ntstc-hun gswe iso d or Kon vivien. - Genetica 11 : 399- 450.

- (1950) : A theory of systemat ics. - Bibl. Biotheorctica 4: 11 7- 180. D ARWIN C. (1859): Tho origin of species b y m eans of ll atural selection. - London. - (1868): The variation of animals and pl a nts und er dornest ication. - L ondon. [) p; CANDOLLE A. P . (1 813): Theori e Clcmo11tairc do la bo taniquo. - P [\,ri s . DnB7.HANRKY T. (l !.l37 ): Gen et ics and the orig in of species. - N ew York . - (195 1) : Genetics and t h e or ig in of sp ec ies . Third edition. - New York. Du RrnTz G. E. (1930): The fundamental units of biological taxonomy. - Svensk. B ot . Tidskr.

24 : 3:33- 428.

Eum-TAHT F. (1788): Botanisch e B em erkungen. - B e iLr. Naturk. vcrw . Wiss . 3: 58- 95.

FHA NDSBN K. J. (1947): Th e experimental formatio n of Brassica napus L. v a r. oleifera DC. and Brassica carinala Braun. - D an sk Bot. Arkiv 12, 7 : 1- 16.

GAJEWSKI vV. (1957): A cytogen ctic study on the genus Geum. - Monogr. Bot. 4 : 1- 4 16.

GOODSPEED T . H. (1954) : The genus Nicotiana. Origins , relationships and evolution of its sp ecies in the li ght of the ir distribution, morphology a.nd cytogonotics. - Chron. Bot. 16 : I - XXII, 1- 536.

GRANT W. F. (1960): The catego1·ios of classical and experimental taxonomy and the species concept. - R evu e Canad. Biol. 19 : 241- 262.

HEYWOOD V. H. (1958) : The presentation of taxonomic information. A short guide for contri­butors to Flora Europaea. - L eicester.

- (1960): The taxonomy of poly ploids in Flora Europaea. - F eddes Repert. 62 : 179- 192. H srnH S. C. & OKA H. I. (1958): Cytological studies of st erility in hybrids b etween distantly

related varieties of rice, Oryza sativa L. - Jap . Jour. Genet. 33 : 73- 80. · HUSKINS C. L. (1931): The origin of Spartina Townsendii. - Genetica 12 : 531- 538.

137

HYLANDI~ H. N. (1945): Nomenkla tori sch e uncl systom a tisch e Studien iibe r nordi sch e Gefoss­pfl an zcn . - Uppsala Univ. Aniskr. 1945, 7 : 1- 337.

Juz 1~rczuK S . V. (1958): K omarovsk aya kontsoptsiy a vida y eyo istorich esk oyc razv itiye i otraz­h eniyo vo " Flore SSSR". - Problema vida v b o tanike I : 130-204.

K1rnN:1<at A. v. (1 866): Gute und schl ochte Arten . -- Inns bruck. Krn 1,1,ANDl~ It C. L. (1953): Inducorl ancl natural p o ly plo id y in P icea Abies (L .) K a rst. - Proc.

VlJ .Lnt. B ot . Cong r . (1050) : 303--304. KIHA RA H. ( l 9fi !)): Consideration s on the orig in of culti vated rice . -- Se iken Zih6 10 : 68- 83. LA MA HC'K J. R. P.A. d e (1809) : Philosophic zoologique . - P aris . L A1\IPHl<: CHT H . (1959): D er Art bogriff, seine Ent wicklung und experimen te ll e Klarl Agung . --­

Agri Hort . Gen et. 17 : 105- 26L1. L AHSEN K . (UlGO) : Cytological a nd exp erimenta l studies on the :fl owerin g plants of the Canary

I s land s . - Biol. Skr. Dan. Viel. Solsk. 11 , 3 : I - GO. LI NNAlWS C. (1751): Philosophia bota nica, in qua ox plicant ur Funda.rnenta bot anica c um d ofini­

t ion ibus p arLium , exemplis t erminorum, o bservationibus r a riorum, a dj octis figu r is aen eis . -S tockholrniae .

LOHAMM AR G. (1931) : Two chromoso me numbe rs in B ntom us nmbellatus L. - Svon sk . Bot. Ticl skr. 2fl : 4!l5- 499.

- ( IH54) : Bulbils in the in:flor0seenees of B utonius1tmbellatus.- Sven sk Bot. Tidskr. 48: 485--488. L bvE A. (HM,3): Cy t ogenoti c studies in Rurnex subgenu s A cetosella. - H e reditaR 30: 1- 136. - (1944a ): Th e cli oec ious fo rrm; of R um ex subgenus Acetosa in Scandinavia . - B ot. Notisor

L ~)44 : 237- 252. - .(1944b): A n ew triplo id B etula verru cosa. - Sven sk Bot. Tid skr. 38: 381- 393. - (1949): Agom ospe rmy in A cetosa. -- H e roclitas 35 : 390- 393. -- ( 19.51): T a xonomical evaluation of p o ly ploids . - Cary ologia 3 : 263- 284. - (1954) : The fo unda tions of cyto taxonomy . - VIII Congr. lnt. B ot., Rapp. & Comm. 9 & 10

59- GG . - (19GOa) : Bi osystemat ics and the processes of speciation. - R oy. Soc . Canad . S tudi a Varia 4:

E volution: i ts sC' ien C'f' a nd d octrine : 11 5- 122. - (1960h ): RpcC' ies con cept a nd tax on om y - a prelud e. -- R evue Canad . Biol. 19: 216- 218. - (1960c ): Biosys te matif's and class ification of apornicts. - F eddes R e port . 62 : 13G- 148. - ( l 960d): T axonomy and chromosom es-a reiteration. - F eddes H ep ert . 62 : 192--202. - (1H61 ): Hylandra - a n ow genus of Cruciforae . - Sven s k. B ot. Tidskr . 55 : 211- 217. L ovE A. & L i)vE D . (1948): Ch rom osom e numbers of Northern plant species. --- lcel. Univ. Ins t.

Appl. Sci ., D ept. Ag ric . R ep . B , 3 : 1- 131. - & - (l 949) : Geo botanical s ig nificance of poly ploidy . I. Poly ploidy and la titude. -- Portug .

Acta Biol. (A), H . B. Goldschmidt Vol.: 272-352. - & - (1956 ): Cy t o taxon omical consp ectus of the Icelandic flora. - ActaHorti Gotob. 20:65- 291. - & - (l957a): Biosyste m a tiquo du genre A corns. - Ann. l'ACFAS 23 : 100. - & - (l957b): Drug cont(~nt and poly ploidy in Acorus. - Proc. Genet . Soc. Can ada 2 : 14- 17. - & - (1961 a ): Some nomen clatura l changes in the European flora. II. Subsp ec ific cat egories. -

B ot. N ot iser. 114 : 48- 56. - & - (196lb): Chrom osome number s of Central and Northwest European plant sp ecies. - ­

Opera B ot a nica 5 : 1- 581. U .ivE D . (1944-) : Cyt ogen etic studios on di oecious ~Melandrium. -- Bot. Notiser 1944 : 120- 213. Lo v E: D. & D ANSKltE AU P . (19 59): Hiosystomatic studios on Xanthium: t axonomic apprai sal and

Cf'o log ical s tatu s . - Can a.cl. J o ur. Bot. 37 : 173- 208. L OVE D . & NA DI:<:AU L. (1961): The Hutchinson polygraph, a m ethod for RimultaneouR expression

of multiple ancl v a riable charac ter s. - Ca nad. Jom. Gen e t. Cyt ol. 3 : 289--294. M AC K EY J. (1954): The t axonom y of h exaploid whoat. - Svonsk B ot. Tidskr. 48 : !179- 590. MANSF I<: LD R. (194-9): Uber d en B egriff d er Art in d er syst em a tisch on Ilota nik. - Biol. Zen­

tralbl. 67 : 320- 331. MARDSDEN-JO NES E. l\1. (1930): The gen eti cs of Gw m interm er/,i wm Willd. h a ud Ehrh ., and its

lw,ck -c· rossoR. - Jour. Gen e t. 23 : :n7- :rn5. MAHS LJ J•;N-.Jn NES E. M . & T1 J1tRILL W. B. (1957): Th e bladder campion s (S ilene ma.rit,ima and

8. 1· 11f!Ja ris ). - R ay Soc . Publ. 139 : 1- 378. MAY!l I•:. ( l !l4-0 ): Spec iation phe nom en a in birds . - Amor. N a turalis t 74: 249- 27 8. - (1942): Systematics and t he o ri g in of species. - N ew York. - (19fi7): Spec ies concepts and d efinitions . - The Sp ecies Problem: Ame r. A ss. Adv . Sci. Publ.

50 : 1- 22. M uNTZI NO A. (1930): Outlines to a gen etic monograph of the genuR Galeop sis. - H creditas

13 : 185 -341. - (1932): Uy to-gen eti c investigations on synthetic Galeopsis 'I'etrahit. - H er editas 16 : 105- 154.

138

- (1937): The effects of chromosomal variation in Dactylis. - H ercditaR 23: 11 3-235. - (1943a): Characteristics of two haploid twins in Dactylis glomerala. - H cr0d ilas 29 : 134- 14-0. - (1943b): Fertility improvement by recombination in autototraplo id Galeo11si::; pubescens.

Heroclitas 29 : 201- 204. - (19f>3): Polyploiderna8 ekologi. - Kg!. Lantbruksakad. Ti lskr . 92 : 119- 130. - ( 1955): Chromosomes in r e lation to spec ies differentiation and plant Lrccd ing . -- Conference

on Chromosomes 1955 : 1- 37. NANNFELDT J. A. (1938): Poa maroccana N::wnf. n. sp . a nd P. rivulorum ]\'faire ot Tribut, two

more tetrnploids of sect . Ochlopoa A. & Gr., a nd some a dditional notes on Ochlopoa. - Svensk. Bot. Ticlskr . il2 : 2%- 321.

NoTWENSKLhLD H. (1945): Cytogenetic stuclies in the genus Phfewn. - Acta Agric . Suocana I , 1 : 1- 138.

- (1!)53): A goneticml study in the mode of segregat io n in h exaploicl Phleum pratense. - H ere­ditas 39 : 4()9--488.

OsTE 1w1mN G. (1 !J42): Chromosome numbers in Anthoxanthum. -- Horeditas 28 : 242- 243 . l'AnmsH M. J. (1!)57): A cytogonetic cornparison of the fl oras of sontheast Asia ancl t h e south­

eas tern Uniterl States. - D iss. Abstr. 17 : 2774- 277G. PIZZO.LONGO P. (1958): R icerch e cariotassi nornicho s u akunc Harnmnclidah~s. - Ann. di Bot. 26: 1- 18. RAY J. (1686): Historia planta rum, sp ecies hactenus editas a li asque irn=mper mnltas noviter in ­

vontas atque dcscriptas cornplccten s . etc. Tonrns p r imus . - L ondini. RoTHMALJ<rn, W. (1!)55): Allgcmci ne Taxonom ie uml Chorologie der Pf-lanzen . - Jena. f{,ozMus M. (1959): Cytological inv0stigat ions on Antho~rx1,nthnm alpimun L. & L ., a n ew sp ec ies

of the flora of Poland. - Acta Biol. Crncov. Ser . Bot. 1 : 171- 184. SARKAR P. & fh• 1"Bnrns G. L. (HJ56): Morph ological evidence co ncPming tlie o ri g in of the B gen­

ome in wheat. - Arner. Jonr. Bot. 43 : 297 --304. SAx K. (1933): Species h ybrid s in Platanus and C'arnpsis . - J our. Arnold Arb. 14: 274- 278 . SEAitS E. H. (1948): The cyto logy and genetics of the w h eats and their re latives . - Adv. in

Ge net. 2 : 239- 270. - ( 1959): The syst0matics, cytology and genetics of whoats. - - H andb. d . Pflanzenziichtung II :

164- 187 . SEMENOV-'l'IAN-SHANSKV A. (1910): Die taxonm:nisch en Grenzen <lor A rt und ihre Unterteilun­

gen. - Berlin. SENDTNER 0 . (1854): Die Vegetations-Verhaltnissc Siidbayerns, nach don Grundsiitzen d er

Pflanzengoographie und mit B cz ug na hrn e auf l.andeskultur geschildc rt. - Munchen. STRRHTNS G. L. (1950): Variations and evolution in plants. - Now York. - (1959): The role of hybridi zation in evolut ion. - Proc. Amer. Philos . Soc. 103 : 231- 251. ST1·~ENJS C. G. G. J. VAN (1957): Specific a nd infraspec ific d elimitation. - Flora Malesia na ,

Ser. I, Vo l. 5,3: CL XVII--CCXXXIV. SwmTLINRKA Z. (1960) : Spontaneou s polyploidization in Rumex hybrids . - Acta Soc. Bot. P olon.

2!J: 79- 98. TrscHLER G. (1951): A llgemeine Pflanzonkaryologie . 2. Halfte : K ernteilung und K ernver ­

sr:h m elznn g. - Hand. d . P fl anzenanatomie II : 1- 1040. --- (1953- 1\)57): A llgem eine Pflanzenkaryolog ie . Erganzm1 ;;sband : Angewandto Pflanzenkaryo-

logie. - H anclb. d. P fl anzenanatorn ie II: 1- 1072. Tum~ssoN G. (1922a): The s pec ies and the variety as ecologica.l u nits. - H ereditas 3 : 100- 113. -- ( 1922b): Tho genoty pieal r espon se of Urn plant sp ecies to the habitat. - Hereditas 3 : 211- 350. VALEN'.rINE D . H. (Hl49): The u nits of experimental taxonomy. - Acta Biotheoretica 9 : 75- 88. v ALENTlNJ'i n . n. & L L)V J" A. ( 1958): T axonomic and biosystematic categori es. - Brittonia 10 :

153--Hiu . W ETTSTEIN R. v. (1 89!l): D er Saison-Dimorphismus als Ausgan gRpunkt fur die Bild ung n cu er

Arton irn Pflanzenreich. - Der. Dtsch. Bot. Ges. 13 : 303- 313 . - (18!J6a): Zur Systemtttik der ouropii isch enEuphrasia-Arten . - Osterr. Bot. Zeitschr. 46: 381- 386. - (1 896b): D ie eu ropii.ischon Arten der Gattun g Gentfona a u s d er Sektion Endotricha Froe l.

uncl ihr ontwicklungsgeschichtli cher Zusamrncnhang. - D enk sch r . Akad. Wiss . Wien, math.­n at. Kl. 04 : 309--382.

- (1898) : Grundziigc der goographisch-morphologisch en Methode cler Pflanzensystematik. - Jena. WrnGE 0. (1917) : The chromosom es . Their number and general im portance. - C. r. Trav. Lab .

Carlsberg 13 : 131- 275. - (19 26) : Artkry d sningsproblemer i Planteriget. - Nordi sk J ordbrngsforskning 8/9: 592- 606. W ULFF H. D. (1940): Uber die Ursache der Sterilitat des Kalmus (Acorus C'alamus L.). - Planta

31 : 478- 491. - (1954): Zur Zytologie, gcographisch en Verbreitung und Morphologie d os K almus. - Arch.

Pharmazie 287/59 : 529- 541.

139