J. Hered. 41(6): 159-163 (1950)

On Somatic Chromosomes of Allium and Tradescantia


*Research Assistant in Cytology, Department of Botany, University of Wisconsin, Madison, Wisconsin; Associate Professor of Botany, and Graduate Assistant, Department of Botany, Michigan State College, East Lansing, Michigan, respectively. The authors are indebted to Professor C. L. Huskins and his associates of the University of Wisconsin for valuable discussion, to Mr. Alfred Owezarzak for the photomicrography and to Professor C. L. Gilly of Michigan State College for critical reading of the manuscript.

SEPARATION of somatic chromosomes (as distinct from chromatids) into two or more groups within a single cell is a phenomenon which is induced by numerous chemicals. Such separations appear to be without benefit of the normal spindle mechanism. A number of terms have been applied to this phenomenon, none of which is very satisfactory. In the present paper we are using the terms separation and reduction or reductional groupings† without reference to the segregation of homologous chromosomes which has been studied in Trillium13 and is also currently the subject of study in other favorable organisms.

The chemicals used may be divided into two groups, albeit somewhat arbitrarily: (a) "mitotic poisons" of which colchicine is the type substance and (b) "physiological substances" of which the sodium salt of nucleic acid may be taken as the type. To the first group may be referred such chemicals as chloral hydrate,11 dimethyl arsenate,10 benzene vapor,2 methyl naphthoquinone12 and ethylene glycol3 as well as colchicine. To the second group belongs salts of nucleic acid and inorganic phosphates.4,5 Perhaps induction by cold and spontaneous occurrence4,6,8,13 should be added to this group.

Such division, whether fundamentally justified or not, serves to emphasize the two major current points of view. These are: (1) that the mechanism is probably the same regardless of the chemical used and, in any event, it is a phenomenon of little theoretical or practical importance,9 and (2) that the mechanisms may be quite different and that there are, in any event, important theoretical and practical implications.6

Our opinion in general is much less precise. We believe that as yet too few comparative studies have been made to warrant any decision as to the uniformity of the mechanism involved. But, we are of the opinion that such departures from "normal" are likely to prove indicative of the kind and operation of "forces" involved in nuclear division and that continued investigation is, therefore, desirable.

Figure 10

Figure 11
A—Fully contracted "metaphase" chromosomes showing peripheral arrangement (Allium). B—Prolongation of peripheral arrangement; separation of chromatids can be seen (Allium). C—A later stage of the same type of arrangement as shown in B (Allium). D and E—"Metaphase separation" (Allium). G, H, and I—Reductional grouping showing separation of chromatids. (Tradescantia and Allium). A to D—Prophase reductional groupings from early prophase (A, B and C) to prometaphase (E), (A, Tradescantia; others Allium). E—An 8-and-8 metaphase separation (Allium). F—A 9-and-7 reductional grouping with chromatids well separated (Allium).

Prior to the publication of Levan and Lotfy's discussion,9 we had noted similarities, with regard to reductional groupings, between colchicine and sodium nucleate treated materials. Preliminary comparative studies on root tips of Tradescantia and Allium have raised a number of questions which we think should be considered rather extensively before any decision as to the relative merits of these views can be justified. It is our present purpose to present the following points for consideration by those interested in the problem:

1. Arrangement of Chromosomes

After colchicine treatment, fully contracted chromosomes frequently show a peripheral arrangement (Figure 10A) usually with the greatest aggregation at two "poles" (Figures 10 D, E). Such peripheral distribution is, of course, normal for prophase but is not usually retained or prolonged (Figure 10 B, C)

No comparable arrangement of fully contracted chromosomes has as yet been found after sodium nucleate treatment. This would appear to indicate a difference in mode of action of the chemicals which, however, may be one of degree rather than kind. Nevertheless, the observational basis for distinction does exist and, until it is superseded by a physiological basis of contrary indication, we are not justified in assuming identity of effect.

2. Frequency of Reductional Groupings

The relative "effectiveness" of the two treatments (colchicine and sodium nucleate) in producing reductional groupings is difficult to assess; both have proved to be highly variable. Sodium nucleate has induced frequencies of reducing nuclei ranging from 2 to 35 percent while, with colchicine, the range has been from 6 to 20 percent. In neither case does duration of treatment appear to be a clear-cut factor, although with colchicine there is some evidence of optima at 6 to 8 and 24 to 36 hours. Likewise in neither case have we been able to find any marked relationship between concentration and effect. However, there is yet relatively little information on this point. The bulk of our comparative data have been derived from treatments with 2 percent sodium nucleate and .01 percent colchicine. Undoubtedly at least three factors must be considered: a) concentration of the chemicals used; b) duration of treatment; and c) rate of nuclear division. Such a consideration will call for much more extensive and complex studies than have so far been made. One difficulty which we foresee is that of expressing the effective concentration of one of the chemicals in terms of the others.

3. Chromosome Distributions

Previous publications6,13 have indicated that there is not only a tendency towards separation of the chromosomes into two equal groups in the case of sodium nucleate induced reductional groupings but that there is also a tendency towards preferential separation of homologues. This tendency, of course, is expressed in relation to the theoretical expectation based on the binomial distribution. There is, pro tem, no definitive hypothesis which would lead us to choose such a distribution as a basis for such expectations. However, this point is discussed by Dr. K. Pateau in a paper now in press. In any event, it is probably significant that colchicine induced separations (Figure 10, G, H, and I) have thus far, been found to show little or no departure from such expectation in distribution of chromosomes. For example, in our current experiments on Allium root tips, 57 out of a total of 173 sodium nucleate induced separations have been found to contain groupings of 8-and-8 whereas expectation is approximately 34. This is a difference which is highly significant (x2 = 15.8, p<0.01). Thirty-five out of 187 colchicine induced separations have been found to contain 8-and-8 groupings where the "expected" number is 37 (x2 = 007, p=ca. 0.9).

4. Types of Groupings

After both treatments, the fully contracted chromosomes may he separated either into two or more groups or spread more or less evenly throughout the cell. Also, in both cases, “unoriented" chromatid separations may occur either with or without accompanying distribution into groups. Likewise, chromatids may clump more or less in the center (“nonpolar"), at one end of the cell ("unipolar"),1 or, in three or more regions ("multi-polar") as in Figure 10F.

The two treatments differ with respect to the time in the mitotic cycle when groupings may occur. Sodium nucleate induced separations are found at all stages of prophase (Figure 11 A-D) as well as later (Figure 11, E-F). No single clear-cut case of prophase groupings has yet been found in colchicine treated material. Only stages of full contraction appear to be affected (Figure 10, D, E, G, H, and I).

5. Prophase to Metaphase* Ratios

*The use of the terms "metaphase" and "anaphase" in cases where there is no true equatorial plate and no organized separation of chromatids is, perhaps, unfortunate. However, since any hastily devised alternatives are likely to be even more misleading we are continuing to use these terms with the explanation that by metaphase we mean a stage of chromosome contraction as great as or greater than typical of normal metaphase and by anaphase, chromatid separation whether orientated or not.

Both sodium nucleate and colchicine appear to reduce the ratio of prophase to metaphase in contrast to untreated materials. Preliminary analyses of current experiments have given ratios which are based on a minimum of 300 cells in each instance (Table I).

The number of prophases is not necessarily reduced; rather the implication may be that onset of "metaphase" is deferred or its duration prolonged by the inhibition of anaphase. Such a phenomenon may or may not have a direct connection with the problem under discussion; but it should, in any event, be taken into consideration.


There is no doubt that a wide variety of chemicals induce separation of somatic chromosomes into two or more groups within one cell. The pertinent question is whether or not the basic mechanism is the same in all cases. To this question we have, as yet, no answer. We have set out, above, a number of bases of observational comparison which we consider to be potentially diagnostic.

TABLE 1.—Ratios based on 300 cells.

  Prophase : Metaphase
Untreated material 4 : 1
Sodium nucleate treated material 2 : 3
Colchicine treated material 1 : 3

The major differences between the effects of colchicine and sodium nucleate seem to be with regard to the frequently observed peripheral arrangement of fully contracted chromosomes after treatment with the former and the relatively high incidence of prophase reductional groupings after the latter. To these may be added what is apparently a difference with respect to the numerical distribution of chromosomes to the groups. Sodium nucleate induced separations show what seems to be a significantly higher proportion of equal groupings than is effected by colchicine, or expected on the basis of the binominal distribution. While this difference appears to be the most striking one, colchicine experiments have not yet been sufficiently extensive to warrant any precise statistical analysis. We may ultimately find less difference than appears to date.

Certainly the proponents of the view' that these separations are basically the same, regardless of the chemical involved, will have to fit such differences as have been indicated herein into any mechanism which may be proposed. Those who favor the view that there are essential differences between, for example colchicine and sodium nucleate induced separations must be prepared to accept some rather striking parallels.

Our opinion is that the answer may well emerge from a comparative study of the differential responses, to various treatments, of the "forces" involved in the spindle mechanism. It is probably pertinent that inorganic phosphates have been shown to delay the breakdown of the nuclear membrane 4 which must almost certainly he correlated with development of the spindle mechanism. A similar though not usually so marked an effect on the membrane has been noted after sodium nucleate treatment but not, in our experience, after colchicine. Careful comparison of responses other than chromosome separations or reductional groupings is therefore indicated.

Literature Cited

  1. ALLEN, NORAH S. Induced mitotic 'Reductional Groupings' in Root Tips of tetraploid Tradescantia. M. S. Thesis, University of Wisconsin. 1949.
  2. BERGEN, C. A., E. R. WITKUS and B. J. SULLIVAN. Bull. Torrey Bot. Club 71:620-623. 1944.
  3. D'AMATO, F. Hereditas. 34:83-103, 1948.
  4. GALINSKY, I. Jour. Hered. 40:289-295. 1942.
  5. HERSHCOPF, W. WEISZ. Induced Reductional Groupings in Root Tips of Tradescantia. M. S. Thesis, University of Wisconsin. 1949.
  6. HUSKINS, C. L. Jour. Hered. 39:311-325. 1948.
  7. ————— and K. C. CHENG. Jour. Hered. 41:13-18. 1950.
  8. ————— and L. M. STEINITZ. Jour. Hered. 39:327-335. 1948.
  9. LEVAN, A., and T. LOTFY. Hereditas. 35:337-373. 1949.
  10. LUDFORD, R. J. Archiv. fur Experimentelle Zellforschung 17-18:411-441. 1936.
  11. NEMEC, B. Jahrbücher. fur Wiss. Botanik. 39:645-730. 1904.
  12. NYBOM, N. and B. KNUTSSOM. Hereditas. 33:220-234. 1947.
  13. WILSON, G. B. and K. C. CHENG. Jour. Hered. 45:3-6. 1949.

Somatic Segregation Biblio