Experiments in Genetics (1925)
Charles Chamberlain Hurst

Chromosomes and characters in Rosa and their significance in the origin of species1

The genus Rosa is admittedly a difficult one in many ways. It has defied the efforts of many systematists since the time of Linnaeus (1753) who remarks in Species Plantarum that the species of Rosa are very difficult to determine, and he adds naïvely that those who have seen few species can distinguish them more easily than those who have examined many. In spite of the strenuous and life-long labours of Lindley, Christ, Crepin, Baker, Almquist and others, it must be admitted that no entirely satisfactory classification of the multitudinous forms of the genus has been made.

Genetics of Rosa

Genetically the genus is almost unknown, and the technical difficulties of raising several generations are very great. My own experiments begun at Burbage 15 years ago are still in their infancy, as the normal time for a generation in many species is about 6 to 8 years, and 6 years were lost owing to the War. The genetical slowness of Rosa material is however fully compensated by its permanence, for the individual plant is able to live to an extreme age, and its life can be indefinitely prolonged by vegetative propagation.

Thanks to the facilities granted by the University Authorities at Cambridge it has been possible for me to work out the chromosome complexes of a number of unexamined forms of Rosa from material collected at Burbage. The results fully confirm the findings of Tackholm (1920), (1922), and Blackburn and Harrison (1921), that the fundamental number of chromosomes in Rosa is 7.

Septuple chromosome numbers

Further the results show that all the somatic and gametic chromosomes examined are either 7 or a multiple of 7. Somatic chromosomes are diploid 14, triploid 21, tetraploid 28, pentaploid 35, hexaploid 42, or octoploid 56. (Fig. 169 a through f)

Tackholm (1922) reports 4 cases of aneuploid Rosae, out of 293 examined, which he presumes are F2 or F3 hybrids, and it is possible that cases of aneuploidy may occur in Rosa species though, so far, I have not been able to find one.

Gametic chromosomes, male and female, are either equal with 7 or 14 or 21 or 28 each, or unequal and matroclinous with a maternal bias in the ratio of 1.5 or 2 or 3 or 4 or 5: 1. All are 7 or a multiple of 7, for if they were not, aneuploids would be frequent (fig. 169 g through l).

Septets of Chromosomes.

So far all the somatic and gametic chromosomes of Rosa appear to be equal in size and alike in form in both diploid and polyploid species. In the polyploid species however the nuclei are correspondingly larger as the number of chromosomes increases.

The significance of the septuple numbers of chromosomes in Rosa is apparent in various stages of gametogenesis and also in some somatic divisions in diploids and polyploids, in which it is evident that the chromosomes are working in sets of seven or septets. (Figs.170 through 172.)

Significance of the Septets.

At the time no special significance was attached to the septets beyond that it confirmed previous observations that in Rosa the chromosomes were working in sevens. Like others one regarded tetraploidy and polyploidy as simply a duplication and reduplication of the chromosomes of the diploid species such as one finds in OEnothera [Gates (1915)], Primula [Gregory (1914), Farmer and Digby (1914)], Datura [Blakeslee (1920)], and which the author has also found in several cultivated forms of Rosa. One looked upon the septets of chromosomes in the spontaneous species of Rosa as sets of seven chromosomes which had in some way been duplicated and reduplicated. But one day when comparing the taxonomic characters of the species in the living collection at Kew, one was struck by the fact that the tetraploid species showed the combined characters of two distinct diploid species, while the hexaploid species showed the combined characters of three distinct diploid species and the octoploid species showed the combined characters of four distinct diploid species.

Five differential diploid species.

These observations suggested that the four double septets in the octoploid species were not simple reduplications of the double septet of a single diploid species, but, on the contrary represented the four differential septets of four distinct diploid species in combination.

Taxonomic analyses of the 400 forms of Rosa examined cytologically by Tackholm (1922) and myself, from material collected at Kew, Paris, Cambridge and Burbage, fully confirmed the previous observations, except that the characters in the polyploid species were found to be represented in five differential diploid species instead of four, which number (five) corresponds with the five septets of chromosomes found in Rosa. So far a decaploid species with five double septets of chromosomes has not been discovered, but four irregular hexaploid forms with five differential septets of chromosomes and characters are in existence (R. Jundzilli Bess., R. glutinosa leioclada Christ, R. stylosa evanida Christ and R. inodora Fries), and it is not at all unlikely that the missing decaploid species may soon be identified.

Whether the missing decaploid be found or not is, however, immaterial since the four forms above clearly demonstrate that five distinct septets of chromosomes and characters may be present in Rosa. That these five septets are distinct from one another is clear since the above forms show the characters of five distinct diploid species.

The five differential septets of chromosomes in Rosa may therefore be distinguished as A, B, C, D, and E and the double septets in differential diploid species as AA, BB, CC, DD and EE.

In view of the remarkable results achieved by Morgan and his colleagues in Drosophila, we may now safely presume that the A septet of chromosomes carries the factors which determine the A set of characters in the AA diploid species, and so on with the B, C, D, and E septets of chromosomes in the BB, CC, DD, and EE diploid species, respectively.

Since the five diploid species AA, BB, CC, DD, and EE carry only one distinct (double) septet each, they may be distinguished as the five differential septet species of Rosa.

So far about 50 different taxonomic characters have been recognized in each of the five differential septet species (fig. 173). Each set of septet characters is distinct from the other four sets and no corresponding septet characters are precisely alike. An analysis of the septet characters shows that while some of them are constant, others are variable and alternative, which leads one to suppose that the constant characters are homozygous while the alternative characters are segregates of heterozygous forms within each septet. This explains the great variability that is found within the five differential septet species, while the general facies remains the same. Some systematists make as many as 40 species out of the five, largely on the ground of their geographical distribution and isolation, but the evidence collected shows that these 40 forms are more reasonably interpreted as geographical sub-species.

It is evident from these and other considerations that the five differential diploid species of Rosa are real discontinuous species, with no intergrading or transitional forms, since each has its own set of at least 50 taxonomic characters which are distinct from the other four sets of characters. Each has a septet of chromosomes which is double in the somatic cells and single in the gametes and presumably the five differential sets of taxonomic characters are represented in the five corresponding differential septets of chromosomes.

Differential septet characters in polyploid species.

The polyploid species of Rosa, so far known, are triploid, tetraploid, pentaploid, hexaploid or octoploid in their septets of chromosomes.

Certain cultivated triploid and tetraploid forms are obviously duplicated forms which have arisen in various ways under cultivation by duplications of the septets of chromosomes in the original diploid species. These naturally maintain the specific characters of the original diploid species while displaying the various varietal segregations peculiar to the species. In certain cases, e.g. the tetraploid form R. odorata Swt. var. Gloire de Dijon, whose septet formula is AAAA, certain peculiarities of structure (e.g. giantism) are present in the duplicated forms which possibly may be due to the chromosomes having been duplicated by longitudinal splitting as in OEnothera gigas (Gates 1915) rather than by latitudinal fragmentation as in Primula kewensis (Digby 1912 and Farmer and Digby 1914).

On the other hand the spontaneous polyploid species of Rosa examined are clearly not duplicated diploids but differential polyploids, with differential septets of chromosomes since they show in their taxonomic characters various combinations of the differential septet characters of the five differential diploid species. For instance the tetraploid sub-species R. gallica L. has the septet formula AACC, the pentaploid R. canina L. var. AABDE, the hexaploid R. Moyesii Hemsl. and Wils. AABBEE and the octoploid R. acicularis Lindl. BBCCDDEE.

The various combinations of the characters of the five differential diploid species AA through EE give the 206 compound septet species in the spontaneous differential polyploids.

The manner of working of the septet factors in the differential polyploid species has not yet been fully ascertained, but there is sufficient evidence to show that as a rule each septet works equally with but independently of the other septets with which it may be associated. So far no cases of blending nor of the general dominance of one septet over another have been found.

In the pentaploid species however, where there is one double septet working with three single septets as in AABDE species (R. canina L var.), the double A septet seems to be working in twice as many characters as in any of the single B, D or E septets.

In the most complex case studied, in the octoploid species BBCCDDEE (R. acicularis Lindl.), the four double septets seem to work more or less in relays in different parts of the plant at different times and seasons, resulting in a periodic predominance of one septet over another in certain parts of the plant, the general result being more or less a mosaic of the four septets of characters arranged end to end or side by side.

Naturally with four double septets working equally and independently in an octoploid species, only about one-fourth of the characters of each septet can be represented at one time. An analysis shows that in a plant of R. acicularis Lindl. carrying four years' growth of surculi, stems, branches and branchlets, about one-half of the characters of each of the four septets B, C, D and E were represented (fig. 174 e and f).

An interesting case was observed the author's experiments in which several plants of the tetraploid species AACC (sub-sp. R. centifolia L.) grown in a greenhouse for a genetical experiment, developed temporarily in the second season the climbing character of the sub-tropical A septet, a feature usually absent when grown under natural conditions

New system of classification of species.

The septet characters in Rosa provide a natural and precise method of classification of the species of the genus. This classification is based on cytological, genetical, and taxonomic characters combined.

On this basis the genus Rosa may be divided into nine sections according to the numbers of somatic septets of chromosomes present in the species.

The nine sections may be sub-divided into 15 sub-sections according to the numbers of gametic septets present in the species.

The 15 sub-sections may be sub-divided into the 211 simple and compound septet species, according to the possible combinations of the five differential septets of chromosomes and characters. So that the presence or absence of any one of the five septets of chromosomes and characters-double or single-determines the septet species.

This gives a definite and uniform classification of the species of Rosa, in which each species in a section or sub-section differs from another simply in a septet of chromosomes and characters. This is probably as close an approximation to a mathematical species as one can hope to get in living organisms.

Workers in genetics and cytology may find such a system of classification useful in other plants and animals, as I have done in Rosa, since according to it a species is no longer a question of opinion, but a definite entity that can be precisely determined by a count of the chromosomes and a reference to the table of septet characters.

A large number of Linnean species correspond and coincide with the septet species, and there are considerable numbers of doubtful species and forms in Rosa whose positions can be ascertained precisely by the septet system of classification outlined above. For instance, the various forms of R. tomentosa Sm. can be distinguished from those of R. mollis Sm. by the presence of an additional septet of chromosomes and characters in the former which is absent in the latter, so that R. tomentosa has five septets of chromosomes, while R. mollis has only four septets. Each septet species, simple or compound, diploid or polyploid, naturally has its own sub-species, varieties and forms which may be indicated if necessary by the addition of indices to the septet formulae.

Ecological distribution

The most marked feature of the five simple septet species is the striking dissimilarity in their general facies.

A gardener with an experienced eye for the habit of a plant would recognize their distinctness at a glance without reference to their taxonomic characters.

Thus, the tender climbing AA septet species is obviously fitted for temperate and luxuriant conditions of life. The spiny small leaved BB septet species is fitted for extremes of drought and heat. The prickly suffruticose CC septet species is fitted for extremes of drought and cold. The reedy stoloniferous DD septet species to extremes of cold and moisture. The cany long noded EE septet species to extremes of heat and moisture.

From the evidence that I have been able to collect concerning the habitats of the five simple septet species it would appear that each is fitted for a certain habitat. Thus the A septet species is strictly confined to temperate and sub-tropical regions, and is usually found in sheltered places. The B septet species is found in dry, sandy and desert places, with extremes of heat. The C septet species grows in dry rocky places often near the sea, with extremes of cold. The D septet species grows for the most part in cold swamps and marshy ground. The E septet species grows in conditions of heavy rainfall and great heat; under cultivation in England, it is tender and susceptible to drought.

A study of the habitats of these species brings one to the geographical distribution of the species of Rosa which is of peculiar interest.

Geographical distribution

The evidence shows that on the whole the five simple septet diploid species have a definite southerly bias, while the compound septet polyploid species have a distinct northerly bias. In fact a general statement might be made that towards the Pole the number of septets of chromosomes and characters increases, while towards the Equator the number decreases (fig. 175). Thus all the octoploid species with four double septets of chromosomes and characters are, so far as known, circumpolar, and do not extend far below the Arctic Circle except at high altitudes. The hexaploid species with three double septets of chromosomes and characters are found further south, and where they extend far are usually at high altitudes. The tetraploid species with two double septets of chromosomes and characters occupy on the whole a middle position, though those with "cold" septets often extend northwards, and those with "warm" septets often extend southwards.

The five simple septet diploid species are for the most part found in southerly or temperate regions, the AA septet species extending to the sub-tropics and even to the tropics at high altitudes. As a rule the "warm" BB septet species is not found north of China and California, nor the "warm" EE septet species north of Central China and California, while the "cold" CC septet species extends to North Japan and Newfoundland, and the "cold" DD septet species to Sweden and South Canada.

The fact that the polyploid species of Rosa with several differential septets of chromosomes and characters are found only in northern latitudes, suggests that their several differential septets may be useful to them in the extreme conditions of life in which they grow and reproduce themselves. In such conditions it might be useful for a species to be able to respond at different times and in divers ways to great extremes of cold and heat, moisture and drought, and darkness and light, so that a polyploid species with several differential septets of chromosomes and characters would be more likely to survive and reproduce itself in Arctic conditions than a simple diploid species with only one septet of characters.

Evolution and origin of species

This remarkable distribution of the species of Rosa leads one to consider the question of their evolution and origin.

The facts of distribution at once suggest two distinct possibilities concerning their evolution: either a line of descent from one primitive decaploid species by successive losses of septets of chromosomes and characters, or a line of ascent from the five simple diploid species by hybridisation and subsequent duplications of septets of chromosomes and characters. The fact that the polyploid species show the combined characters of the five diploid species seems to support the idea of origin by hybridisation and duplication. On the other hand such a view presents serious cytological and geographical difficulties.

*CybeRose note: Song , K. et al. (1995) Rapid genome change in synthetic polyploids of Brassica and its implications for polyploid evolution. Proc. Natl. Acad. Sci. U. S. A. 92: 7719-7723.
Implications of Rapid Genome Change for Polyploid Evolution. Using synthetic polyploids, we have demonstrated that extensive genome change can occur in the early generations of Brassica polyploids. Genetic diversity accumulated among self-fertilized progenies, even when the starting materials were completely homozygous. We do not know whether these types of changes or this extent of change has occurred in the early generations of natural Brassica or other polyploid species. However, our molecular results, when combined with variation in fertility and other morphological traits observed in our synthetic polyploids and in previous studies, suggest that rapid genome change in newly formed polyploids can produce many novel genotypes that would provide new genetic variation for selection. Thus, rapid genome change could accelerate evolutionary processes among progenies of newly formed polyploids, and this may, in part, account for the success and diversification of many ancient polyploid lineages in both plants and animals.

The cytological difficulty is that two diploid species AA and BB hybridised will not give a tetraploid species AABB in F1 but only a sterile diploid hybrid AB. This hybrid might be fertile by the segregation of whole septets A and B in the male and female reduction divisions, when it would give either sterile diploid hybrids AB like itself, or fertile diploid species AA and BB like its parents. Apparently the only practical possibility of this sterile diploid hybrid AB producing a fertile tetraploid species AABB would be by a duplication of its somatic chromosomes, giving rise to a bud-mutation, which, by self fertilisation, would reproduce its kind. This, however, would give a homozygous tetraploid species*, while all the tetraploid species of Rosa are certainly heterozygous. The possibility of a simultaneous duplication of the chromosomes in the uniting male and female gametes is too remote to be seriously considered. In view of the many repeated duplications of this kind necessary to make the ten tetraploid species and the more numerous polyploid species, the probability of origin by hybridisation on the whole seems to be extremely small, though not altogether excluded. The geographical difficulty of the origin of the polyploid species by hybridisation is that the diploid species necessary to form the tetraploid species are not there to be hybridised. All the five diploid species are widely spread, highly specialised, occupy distinct habitats and usually flower at different times, so that the probability of their being hybridised in the precise pairs necessary to form the ten tetraploid species and the more numerous polyploid species, seen to be to very remote. Even if an AB diploid hybrid did happen to arise in an AA habitat, and succeeded in making itself chromosome septets to a tetraploid species AABB, it is exceedingly doubtful whether it would survive and reproduce itself in the AA habitat long enough to be able to migrate into the AABB habitat farther north or at a much higher altitude.

Further, an evolution of the Arctic polyploid species by hybridisation of sub-tropical diploid species would seem to involve a complete reversal of the generally accepted views of the Arctic origin of the flora of the Northern Hemisphere.

Finally, an origin from five primitive diploid species implies a multiple origin with its consequent ulterior problems, while an origin from one primitive polyploid species is a more simple proposition.

On the whole it seems more easy to suppose that a northern decaploid was the primitive species from which the existing species of Rosa have descended, by successive losses of septets of chromosomes and characters, thus automatically fitting the lower polyploid species and the simple diploid species for their peculiar habitats.

1It is significant that a form of this tetraploid species AACC has produced several triploid forms AAC under cultivation in Holland and France, and it may be significant that the individual form examined by me is said to have been introduced from Persia, where the closely allied pentaploid species AACCE was found growing on Omar Khayyam's grave.

As the conditions of life became less extreme the necessity of the several septets might become less, and by successive losses of septets the decaploid species would give rise to an octoploid species, the octoploid to a hexaploid species, the hexaploid to a tetraploid species until finally the simple diploid species would emerge better fitted for the more temperate and specialised conditions of life in the South. Further it is probable that the conditions of life themselves may have been the direct cause of these speciations in Rosa, for my cytological observations show that a whole septet of chromosomes may be lost, if retarded and left behind in a cell-division, in the species AACC (sub-sp. R. damascena L.) that has been subject to changed conditions of life for a considerable period of time1 (fig. 170 j). (See also fig. 171.)

If, owing to changed conditions of life, the working of a certain septet of chromosomes in a polyploid species was not necessary to the life of the plant, by constant disuse it might weaken, lag behind in a cell-division and be ultimately lost. In such cases new species would arise with a septet of chromosomes and characters less than their parents, and these new species would automatically be better fitted to survive and reproduce themselves in the changed conditions of life.

Irregular polyploid species.

With regard to the origin of the irregular polyploid species of Rosa peculiar to Europe and Western Asia, which consist of tetraploid, pentaploid and hexaploid species with some single septets of chromosomes (instead of all double septets as in the regular polyploid species of Europe, Asia, and America), Tackholm (1920) (1922) and Blackburn and Harrison (1921) agree in concluding that all have originated by hybridisation. It must be admitted that both the cytological evidence and the septet formulae of these species agree with the hybridisation hypothesis. The chief difficulty however in accepting this explanation of the origin of these irregular species is their geographical and ecological distribution, for the necessary parents are seldom there to be hybridised.

Tackholm (1922) explains this by presuming a suitable distribution of the necessary parent species before the Ice Age, but so far as I can ascertain, there is no evidence of this in the palaeontological records. On the whole it seems more probable that the irregular polyploid species of Europe and Western Asia have originated in the same way as the regular polyploid species of Europe, Asia and America by losses of septets of chromosomes, the only difference being that while the regular species have lost a double septet at a time, the irregular species have lost only a single septet of chromosomes and characters. At the same time one must admit the probability that some of the irregular polyploid species, particularly those whose putative parents grow in close proximity, have arisen by hybridisation, while others have arisen from somatic or gametic speciations.

Conclusion.

In conclusion it may be useful to point out the value of the septet scheme of chromosomes and characters in Rosa as a working hypothesis. First, it has already cleared up many of the difficulties pointed out by Linnaeus, Lindley, Christ, Crépin, Tackholm and Harrison in their studies of the genus, and thus enables one to deal with certain problems of classification and origin which before seemed impossible to solve.

It provides a satisfactory explanation of the extraordinary polymorphism in Rosa and goes far to explain the chief canses of fertility and sterility in species and hybrids.

It demonstrates the existence of two distinct forms of polyploidy, Duplicated and Differential.

It gives one a new conception of heredity in the differential polyploid species leading to broader conceptions of mutation, evolution and origin of Species. Thus it shows that the ordinary Mendelian ratios may only be expected in crosses within a diploid species and consequently explains how non-Mendelian ratios may be expected in crosses within a polyploid species.

It provides a new system of classification of species, based on cytological, genetical and taxonomic characters combined, which not only gives one a precise and uniform definition of a species in what is probably the most polymorphic genus of plants and animals, but has the distinct advantage that each species can be verified experimentally by both cytological and genetical methods.

The septet of chromosomes provides a simple yet vital mechanism for the evolution of the Linnean species of Rosa, demonstrating that variations may be large, discontinuous, truly specific and adaptive in the polyploid species (i.e. Speciation), while in the diploid species variations will be comparatively small, sub-specific, varietal and Mendelian (i.e. Mutation).

The septet scheme of speciation by phylogenetic descent from a primitive polyploid species goes far to explain the geographical distribution of the existing species of Rosa in the Northern Hemisphere, and illustrates the point of view that the evolution of species may lead to simplification and specialisation rather than to complexity and co-operation.

Further, it suggests that the existing species of Rosa were largely predetermined in the primitive polyploid species and that their evolution has been more or less automatic in response to the conditions of life which have directly caused and controlled the speciations through the losses of the differential septets of chromosomes by disuse, and in this way the changing conditions of life have determined the trend of the evolution of the species throughout the Northern Hemisphere.

Finally, the septet scheme of chromosomes and characters in Rosa can be verified experimentally by both cytological and genetical methods and the following seven crucial experimental tests should be decisive.

Experimental tests of the septet scheme of chromosomes and characters in Rosa.

(1) Cytological Test. In any species of Rosa, the chromosome number should be found to coincide with the septet formula of the taxonomic characters.

(2) Genetical Tests. All crosses between sub-species or varieties of the same diploid species should breed true to the septet characters of that species and their somatic chromosomes should be diploid (14).

(3) All hybrids between diploid species with different septets should have the septet characters of the corresponding tetraploid species, but their chromosomes should be diploid (14).

(4) All hybrids between regular tetraploid species with different septets should have the septet characters of the corresponding octoploid species, but their chromosomes should be tetraploid (28).

(5) All hybrids between regular tetraploid and diploid species with different septets should have the septet characters of the corresponding hexaploid species, but their chromosomes should be triploid (21).

(6) All hybrids between regular hexaploid and tetraploid species with different septets should have the septet characters of the decaploid species, but their chromosomes should be pentaploid (35).

(7) All hybrids between regular octoploid and diploid species with different septets should have the septet characters of the decaploid species, but their chromosomes should be pentaploid (35).

(The cytological test (1) is obviously more rapid than the genetical tests (2) through (7), and has already been carried out successfully many times. The genetical tests have been put in hand, but naturally some time must elapse before they can be completed.)

Note added 1925.

Full details of the working of the septet scheme of chromosomes and characters in Rosa with numerous photographs, specific tables and maps of distribution already prepared, will be published in the form of a monograph of the genus in several volumes.

Vol. I dealing with "The Five Diploid Species of Rosa" will be issued as soon as possible. [CybeRose note: It didn't happen]

Owing to the considerable time that must elapse before the whole of the detailed evidence can be presented, the above "Summary of Conclusions" has been published in the hope that workers in other genera may be able to test how far the Rosa principles apply to other plants and animals.

List of authors.

FIG. 169. SOMATIC AND GAMETIC CHROMOSOMES IN ROSA.

a. Somatic chromosomes of R. indica L. Diploid 14.
b. Somatic chromosomes of R. provincialis Ait. Triploid 21.
c. Somatic chromosomes of R. gallica L. Tetraploid 28.
d. Somatic chromosomes of R. canina L. Pentaploid 35.
e. Somatic chromosomes of R. Moyesii Hemsl. and Wils. Hexaploid 42.
f. Somatic chromosomes of R. Hilliana Hurst. Octoploid 56.
g. Equal Gallletic chromosomes of R. indica L. Diploid 7 (egg) + 7 (pollen) Diakinesis with 7 pairs of chromosomes.
h. Equal Gametic chromosomes of R.Moyesii Hemsl. and Wils. Hexaploid 21 (Egg) +21 (pollen) Pollen-grain with 21 chromosomes.
i. Equal Gametic chromosomes of R. Hilliana Hurst. Octoploid 28 (egg) +28 (pollen) Diakinesis with 28 pairs of chromosomes.
j. Unequal Gametic chromosomes of x R. Goethe Hort. Triploid 14 (egg) +7 (pollen), Interkinesis with 14 micropylar and 7 chalazal chromosomes.
k. Unequal Gametic chromosomes of R. pomifera Herrm. Tetraploid 21 (egg) + 7 (pollen) Pollen-grain divisions with 7 chromosomes.
l. Unequal Gametic chromosomes of R. Froebelli (Christ). Pentaploid 28 (egg) + 7 (pollen) Late Diakinesis with 7 pairs and 21 single chromosomes.

FIG. 170. CHROMOSOMES IN ROSA, WORKING IN SEPTETS.

a. Pre-diakinesis with 7 pairs of chromosomes in the diploid R. indica L.
b. Later stage, the chromosomes with linin connections.
c. Pre-diakinesis with 7 pairs of chromosomes in the diploid R. Willmottiae Hemsl.
d. Later stage, the chromosomes with linin connections.
e. Pre-diakinesis with 7 pairs of chromosomes in three stages, in the diploid R. indica L.
f. Pollen-grains with 7 chromosomes in male nuclei (dividing) in the diploid R. rugosa Thunb.
g. Diakinesis with 7 pairs of chromosomes (cut cell) and 7 retarded single chromosomes with linin connections in the triploid R. semperflorens Curt.
h. Later stage with 7 pairs and 7 single chromosomes (cut cell).
i. Heterotype metaphase with 7 pairs on equatorial plate and 7 single chromosomes at the micropylar end in the triploid x R. Goethe Hort.
j. Somatic telophase with 7 chromosomes lagging on the equatorial plate after division in the tetraploid R. damascena L.
k. Heterotype metaphase (polar view) with 7 pairs in centre and 21 chromosomes grouped around in the pentaploid R. Froebelli (Christ).
l. Heterotype early telophase with paired septets reduced and 21 single chromosomes lagging on the equatorial plate in the pentaploid R. Froebelii (Christ).
m. Later stage, the reduced septets distinguished by their more advanced condition (in this pentaploid the single chromosomes do not split).

FIG 171. CASES OF THE LOSS OF A SEPTET IN SOMATIC CELLS OF ROSA.

(a) Somatic prophase with 21 chromosomes in the (CDD) triploid in which a septet of chromosomes has been excluded from the nucleus.
(b) A later stage in the same triploid.
(c) Somatic plate in the same triploid with a septet of chromosomes detached from the others.
(d) Somatic telophase in the same triploid in which a septet of chromosomes has been excluded from the division and appears to be degenerating in the cytoplasm.
(e) Somatic prophase with 21 chromosomes in the (AAA) triploid R. semperflorens Curt. in which a septet of chromosomes is being excluded from the nucleus.
(f) Somatic plate in the same triploid with a septet of chromosomes detached from the others.

FIG. 172. CHROMOSOMES IN ROSA WORKING IN SEPTETS IN EMBRYO-SAC
MOTHER-CELLS OF IRREGULAR POLYPLOID SPECIES, SHOWING UNEQUAL REDUCTION.

(a) Heterotype metaphase with 14 pairs of chromosomes (two double septets AACC) on the equatorial plate (2 pairs already reduced), and 14 single chromosomes (two single septets BD) towards the micropylar pole, in the irregular hexaplaid (AABCCD).
(b) Heterotype telophase in the (AACCDE) irregular hexaploid R. alba L. with 28 chromosomes at the micropylar pole (ACDE) and 14 at the chalazal pole (AC), leading to the formation of egg gametes with 28 chromosomes (ACDE) instead of 21 as in the regular hexaploids.
(c) Heterotype metaphase with 7 pairs of chromosomes (one double septet EE) on the equatorial plate and 21 single chromosomes (three single septets ACD) at the micropylar pole (cut cell) in the (ACDEE) pentaploid R. Froebelii (Christ).
(d) Heterotype telophase in the same pentaploid with 28 chromosomes (ACDE) at the micropylar pole (cut cell) and 7 (E) at the chalazal pole, leading to the formation of egg gametes with 28 chromosomes (ACDE).
(e) Interkinesis in the same pentaploid with 28 chromosomes (ACDE) in the micropylar nucleus and 7 (E) in the chalazal nucleus (two cuts).
(f) Homotype division in the pentaploid (AABCD), the micropylar cell in metaphase with 28 chromosomes (ABCD) splitting (cut cell), the chalazal cell in telophase with 7 chromosomes (A) at each pole. The above figures are reductions of camera lucida drawings made with Zeiss apochromatic objective 1.5 mm and Zeiss 18 compensating ocular, using tube-length 175 mm. and Watson's Holoscopic Oil-immersion Condenser with aperture 1.34.

FIG. 173. THE FIVE DIPLOID SPECIES OF ROSA SHOWING DIFFERENTIAL SEPTET CHARACTERS (AA through EE).

AA Species (a) R. indica L (b) R. Brunonii Lindl. (c) R. multiflora Thunb. (d) R. moschata Mill. (a) Flowers, (b) Fruit (c) Habit of Branching (d) Surculus with leaf
BB Species (e) (h) R. Willmottiae Hemsl. (f) R. sericea Lindl. (g) R. Hugonis Hemsl. (e) Flowers (f ) Fruit (g) Habit of Branching (h) Surculus with leaf
CC Species (i through 1) R. rugosa Thunb. (i) Flowers (j) Fruit (k) Habit of Branching (1) Surculus with leaf
DD Species (m) R. fraxinifolia Lindl. (n) (p) R. pisocarpa A. Gray. (m) Flowers (n) Fruit (o) Habit of Branching (p) Surculus with leaf
EE Species (q through t) R. macrophylla Lindl. (q) Flowers (r) Fruit (s) Habit of Branching (t) Surculus with leaf

FIG. 174. DIFFERENTIAL POLYPLOID SPECIES OF ROSA WITH THEIR SEPTET FORMULAE.

(a) Tetraploid BBCC (R. spinosissima L.).
(b) Tetraploid BBDD (R. altaica Willd.).
(c) Hexaploid AADDEE (R. nutkana Presl).
(d) Hexaploid AABBEE (R. Moyesii Hemsl. and Wils.).
(e) Octoploid BBCCDDEE (R. acicularis Lindl.).
(f) Octoploid BBCCDDEE (R. acicularis Lindl.).

FIG. 175 GEOGRAPHICAL DISTRIBUTION OF THE DIPLOID POLYPLOID SPECIES OF ROSA.

The figure 2 on the map represents diploid species
The figure 4 on the map represents tetraploid species
The figure 5 on the map represents pentaploid species
The figure 6 on the map represents hexaploid species
The figure 8 on the map represents octoploid species

Hurst Bibliography