Hereditas (1940) 26:277-291






(With a Summary in English)

Several years ago I (LAMPRECHT, 1934) published a work on the heredity of partial color in the testa of Phaseolus vulgaris. This work summarized and discussed the known results in the field at that time, and went on to describe the segregation results in the F2 of seven crosses. The discussion and results can be briefly summarized as follows.

The primary requirement for the formation of partial color is recessivity in the two fundamental genes T and E. All TE plants are wholly colored. In addition, there are several genes for partial color, which should work individually and together to determine the coloration of various areas of the testa. Complex interactions exist between those genes. The color of seeds that are recessive in all partial color genes is still unknown. Yet it is imaginable that the fundamental genes themselves might already determine a certain type of partial color.

Seventeen considerably different types of partial color were described in the work mentioned above. The denominations to be introduced in 1.c. were given Latin names that described the shape of the patterns on the seed surface. For example, bipunctata = with two small spots, virgata = with stripes, arcus = with an arch, virgarcus = with stripes and arch, sellatus = saddle, etc. (see the illustrations to follow below). All in all, about 22 different types of partial color were already known at the time, leading to the assumption that four or five different genes exist for partial color.

The segregation in the crosses mentioned above was unclear, with the exception of cross number 32. The resulting segregation numbers for the different partial color types could not be explained using the well-known simpler schemes for 2 or 3 gene segregation. There were two major causes for this. First, all crosses were carried out between wholly colored and partially colored lines, and it became apparent that in terms of their genotypical constitution for partial color, most of the wholly colored lines were different in several genes. Second, the classification of the partial color types was somewhat uncertain. While there were hardly completely continuous transitions between the less colored types (lying between unipunctata and virgarcus), nonetheless some of those were weakly marked in some part of their partial color. At the time, this was interpreted as having been caused by modifying influences. As shown below, these phenomena mainly stem from genetic heterozygosity for partial color.

So far, only two gene pairs for partial color have been thoroughly characterized in their effects. The first gene pair is Z-z, and applies only to the three types sellatus, ZZ, Piebald, Zz, and virgarcus, zz. Evidence for this segregation comes from E. V. TSCHERMAK (1912), SURFACE (1916), and also from SAX and MCPHEE (1923). The second gene pair was proposed by the author (LAMPRECHT, 1934) and designated Bip-bip, derived form bipunctata. It is one of the genes (see Fig. 1 & 2, below) responsible for the segregation virgarcus (dominant) : bipunctata (recessive) (with intermediate types).

Since then, F. SCHREIBER (1934) has published work according to which there exists a dominant "erasure factor" for partial color. This factor should entirely prevent the formation of partial color, so that pure white seeds result in place of partially colored seeds. I have observed the segregation of white seed plants from partial color plants myself in several crosses (See 1.c.), but without having found a gene that is generally dominant over partial color.


This cross, number 253, was carried out between line 5 of the bipunctata type (derived from the French bean variety Incomparable) and line 88 of the virgata type (selected from the English bean variety Early Giant). Figure 1 shows the bipunctata type (left) and the virgata type (right). Both types exhibit a certain relatively small variation (see Figure 9 in LAMPRECHT, 1934, in which three seeds illustrate the breadth of variation of line 5; see also Figure 19 op. cit., in which the middle and right seeds correspond to the virgata type of line 88).

The seeds obtained from the F1 plants exhibited the patterns on the middle seed in Figure 1. Like its parent line, this seed has the two spots characteristic of the bipunctata type, with the same amount of variation. Not so with the stripe on the micropyle side of the virgata seed. This stripe is no longer completely formed, but is indicated by dots, although these occasionally more or less run together. Among the F1 seeds it was possible to find some that were difficult to tell apart from weakly marked virgata of the parent line.

Fig. 1 A seed of the bipunctata type (left) and of the virgata type, bip Arc (right). The middle seed corresponds to the heterozygote constitution bipbip Arcarc (F1 of cross number 253).

In the second generation, 600 seeds were planted to produce a total of 537 mature plants. The classification of the bipunctata seeds caused no difficulties; they were easy to pick out from the others. With the virgata seeds, however, it seemed difficult to draw a clear border between fully developed virgata types (= parent line 88) and those with stripes composed only of dots (cf. LAMPRECHT, 1934, Figure 19, left and middle seed). Thus, the border drawn between those last two types is not precise. The following numbers resulted for the three F2 types:

Observed: 132 bipunctata : 294 weak virgata : 111 typical virgata
Expected: 134.25 : 268.50 : 134.25
D/m for 1:2:1 = 0.22 2.20 2.31

There is no doubt that this is a monohybrid segregation after the Zea type, but the separation between the heterozygote and homozygote virgata is not entirely clear. In this area there are intermediate types which can only be definitely assigned after inspection of a further generation. The F3 generation confirmed this. A total of 22 families were tested. All the bipunctata types, 6 families with 109 individuals, produced only further bipunctata types. The same was found to be true of the progeny of the typical virgata. Five families with 92 plants remained consistently virgata. Of the 11 families of weak virgata, however, 2 were consistently virgata, now marked with a clear virgata stripe, while the other 9 families segregated in a ratio of 51 bipunctata : 92 weak virgata : 34 typical virgata. This is obviously the same 1 : 2 : 1 ratio as in F2.

Fig. 2 Three seeds from the virgarcus type of line 57 of the variety Gold Rain. Formula: Bip Arc. The left seed has an unusually weakly marked arcus.

Thus, the difference between virgata and bipunctata is attributable to a gene that causes the virgata pattern in its dominant form and the bipunctata pattern in its recessive form. A symbol for this gene will be introduced in the discussion of the next cross.


This cross, number 251, was carried out between line 5 (as in the last cross) and line 57, from the German wax bean variety Gold Rain, representing the virgarcus type. The seeds of the latter show the pattern and variation reproduced in Figure 2.

The seeds obtained from F1 showed the seed pattern on the right in Figure 4. The formation of the virgata stripe, as well as of the arch extending toward both sides of the hilum from the upper spot, varied in intensity but never reached that of the parent line 57 (Figure 2). Sometimes those two patterns were weak, but they were always still recognizable.

Fig. 3 At left, arcus, homozygote and well-developed: BipBip arcarc; at right, the heterozygote arcus type: Bipbip arcarc.

Fig. 4 Three different heterozygote virgarcus types. Heterozygote virgata: BipBip Arcarc (left); heterozygote arcus: Bipbip ArcArc (middle); heterozygote virgata and arcus: Bipbip Arcarc.

The second generation consisted of 552 individuals (from 600 seeds sowed). The resulting seeds were divided into nine different groups according to the nine different possible gene combinations of a bifactoral segregation. The color distributions on the testa and their corresponding genetic constitutions can be seen in Figures 1 - 4.

One of the two genes at work here is the previously recognized gene Bip (LAMPRECHT, 1934). The second I have called Arc, derived from arcus = arch, Figure 3 (left). This so-called arcus type, which is doubly recessive in arc, is thus characterized by a colored arch to both sides of the hilum. The following relationships result from these two genes and the partial color types:

BipBip ArcArc = homozygote virgarcus type, arch and stripe strongly marked; Figure 2.
BipBip Arcarc = heterozygote virgarcus type, well-developed arch, stripe only indicated with more or less clear dots; Figure 4, at left.
BipBip arcarc = homozygote arcus type, clearly visible arch, missing stripe; Figure 3, at left
Bipbip ArcArc = Bip-heterozygote virgarcus type, strong stripe, arch only indicated with dots; Figure 4, in middle
Bipbip Arcarc = Arc- and Bip-heterozygote virgarcus type, both arch and stripe only indicated with more or less clear dots; Figure 4, at right.
Bipbip arcarc = Bip-heterozygote arcus type, arch only indicated by more or less clear dots (sometimes missing completely), stripe always missing; Figure 3, at right.
bipbip ArcArc = homozygote virgata type, arch missing, strongly developed stripe; Figure 1, at right
bipbip Arcarc = Arc-heterozygote virgata type, arch missing, stripe only indicated with more or less visible dots; Figure 1, in middle.
bipbip arcarc = homozygote bipunctata type, arch and stripe missing, only the two spots at caruncula and micropyle are present; Figure 1, at left.

The figures also make it apparent that the bipunctata and arcus types can also exhibit a few dots of color beginning a stripe, but that in these types the coloration is only present in the immediate vicinity of the caruncula and the micropyle (cf. Figure 1, left, and 3). This precludes any confusion with the virgata type (cf. Figure 1, middle, and 4, left and right).

The distribution of the 552 plants of the F2 generation into the groups listed above occurred as follows:

Observed: Theoretically
BipBip ArcArc consistent virgarcus



BipBip Arcarc virgata segregation



Bipbip ArcArc arcus segregation



Bipbip Arcarc virgata & arcus segregation





BipBip arcarc consistent arcus



Bipbip arcarc arcus segregation





bipbip ArcArc consistent virgata



bipbip Arcarc virgata segregation





bipbip arcarc consistent bipunctata



Examination of these numbers can lead to the following conclusions. The individual sums observed for the two groups virgarcus and virgata are in agreement with the expected numbers for class segregations. In the arcus group, however, there is a strong deficit, and in the bipunctata group there is a corresponding excess. If those two groups are combined, then the following numbers result:

Observed: 334 virgarcus: 107 virgata: 111 arcus-bipunctata
Expected: 310.5 103.5 138.0
D/m for 9:3:4 2.02 0.49 2.96

Without a doubt, there is bifactoral segregation for the two gene pairs Bip-bip and Arc-Arc, but the definition of the different heterozygote groups is still unsatisfactory. Within each of the two groups virgarcus and virgata we again find the same uncertainty about the borders of the homozygote, as opposed to the heterozygote, stripes as in the previous cross, but to an even greater degree. Furthermore, the arcus is usually badly developed.

Only 2 individuals were classified as homozygote arcus, and 33 as heterozygote. The remaining individuals, about 60, were in the bipunctata group. This was proven by the third generation, in which both hetero- and homozygote types segregated from bipunctata. However, since the arcus was sometimes well formed in both lines and in other crosses, it seems possible that there is a gene influencing cross number 251 that controls the weak and uncertain formation of this type. This is probably a partial color gene that is contained in both parent lines in an allele form.


This cross, number 174, was carried out between line 5 (as in the previous cross) and line 6 from the French bean variety Tres nain précoce. The latter represents the major type; see Figure 5. The seeds resulting from the first generation display the patterns shown in Figure 6. As can be seen, these seeds are close to the virgarcus type, but with a somewhat stronger development of the arcus and of two patches to the sides of the micropyle. The two previous crosses showed that intermediate types resulted when the Arc and Bip genes for partial color were heterozygote. The same seems to be the case here, but with a stronger tendency toward the type with less color on the testa. Thus, greater spread of color, such as that displayed by one of the parents (line 6), seems to be controlled by recessiveness in a further partial color gene.

Table 1 The segregation of the F1 hybrid Arcarc Bipbip Diffdiff in F2 of cross number 174.

Genetic segregation No. of individuals D/m
Partly colored
Observed Expected
F1 :{ 48 Arc{ 36 Bip{ 27 Diff
9 diff
12 bip{ 9 Diff
3 diff
16 arc{ 12 Bip Diff

155 137 1.77
4 bip Diff

In the second generation, 600 seeds were sowed, from which 548 seed-bearing plants developed. With regard to partial color, those could be divided into the following 6 different types: bipunctata, arcus, virgata, virgarcus, maximus, and major (cf. Figure 1-5 & 7). The individual counts observed and theoretically expected for those six types are collected in Table 1.

Fig. 5 Three sides of the major type from line 6, Tres nain précoce. One of the parents of cross number 174.

Fig. 6 The pattern on the F1 of cross number 174; formula: Bipbip Arcarc Diffdiff.

Above all, it can be seen from this table that a segregation has taken place in the two genes Arc and Bip dealt with in the previous cross, as well as in a further gene, which was called Diff. Since one of the parent lines (number 5) of the bipunctata type is recessive in Arc and Bip, the segregation we have observed leads to the conclusion that the major type represented by the other parent line (number 6) must be dominant in those two genes. Furthermore, the major type must differ from the bipunctata type in at least one further gene. According to the segregation results, that new gene determines the spread of color on the testa. As such it was given the symbol diff, derived from diffundere = spreading.

Fig. 7 Three seeds from the maximus type.

For the two gene pairs Arc-arc and Bip-bip, we observe the same segregation as in the previous cross, corresponding to a ratio of 9 : 3 : 4, since the boundary between the two types arcus and bipunctata is not clear here either. As the arcus is often weakly formed or no longer recognizable, a part of the arcus types become classified as bipunctata. The following numbers result for the segregation of those two genes:

Observed: 304 Arc Bip : 89 Arc bip : 155 arc (Bip or bip)
Expected: 308.25 102.75 137.00
D/m for 9:3:4 0.37 1.51 1.77

As can be seen, the agreement between observed and theoretically expected numbers is satisfactory. Of the arc individuals, 72 were classified as arcus and 83 as bipunctata, which can be traced to the weak formation of the arcus mentioned previously.

The new gene diff does not, at least in the genotypical constitutions that come into consideration here, seem to possess an influence, in either dominant or recessive form. In the group of Arc individuals, recessiveness in diff leads to a strong spreading of partial color. The virgarcus type is transformed into the maximus type by diff, while the virgata type is transformed into the major type. Here we can observe an uncommon situation in which certain genes for a property (partial color) amplify it in their dominant form, while another gene (diff) brings this about in recessive form. That is, Arc and Bip in dominant form cause greater spreading of color on the testa, while Diff causes greater spreading when recessive.

As we have already mentioned for F1, virgarcus seeds show far less spreading of color when Diff is heterozygote (see Figure 6). The same seems to be true for the virgata type. Heterozygosity in Arc and Bip can also be determined from the patterns of seeds of the maximus and major types, which are both homozygote recessive in diff. Figure 8 will serve as an example for this.

Fig. 8 Seeds of major and maximus type from segregation in cross number 174, heterozygote in one or more genes. At left, a major type, heterozygote in virgata (Bipbip) and perhaps also in arcus; in the middle, a seed that is homozygote virgata, but probably heterozygote arcus. At right, a maximus seed with heterozygote virgata and probably also arcus.

As can be seen from the figure, the spreading of color is clearly reduced when Bip is heterozygote. As the left and right seeds in the figure show, heterozygote Arc, which dissolves the stripe of the virgata type into spots, is easily determined. On the middle seed, the strong, broad stripe continues around the end of the seed; here it is clearly homozygote, ArcArc. The patches to either side of the micropyle, brought out by the gene diff, seem consistently clearly marked (for example, compare Figure 8, right seed).

It is fairly difficult to determine a clear boundary between the different heterozygote types, and this could probably only be carried out accurately given a segregation in only one of the genes involved. To this end, further analyses in F3, F4, etc., in crosses specially designed for this purpose, are necessary.


For this cross, line number 57, from the German wax bean variety Gold Rain, was used as the virgarcus type, just as it was a parent in cross number 251. The second parent, line number 106, is probably the result of a spontaneous cross and was discovered as a rogue [anomaly] in a seed sampling.

Fig. 9 Three seeds of the minor type, line number 106, one of the parents of cross number 282.

The multiplication of its progeny showed that these were consistent with regard to partial color. The color of the testa is chestnut brown, with the formula: P C J G b V r t. The minor type (cf. Figure 9) takes on an intermediate position between the major and the minimus types (cf. Figure 10).

Fig. 10 Three seeds of the minimus type, line number 63.

Without having verified the genetic consistency of the minor type, one would be most inclined to regard it as a modification of the major type in which the color has spread somewhat further. In its spreading of color, it is somewhat nearer to the major type than to the minimus type.

The seeds obtained in F1 of this cross showed the patterns reproduced in Figure 11. In terms of their

Fig. 11 Seed obtained from the F1 of cross number 282. Formula: Bipbip ArcArc Diffdiff Expexp.

color spread, these corresponded most closely to the major type, but differed from it by the light cuts directed toward the micropyle to either side of the stripe (virgata). The major type is also missing the light cuts visible here above the patches. Since the stripe (virgata) is fully and strongly developed, the seeds should be ArcArc. Accordingly, we would expect one of the parents, line number 106, the minor type, to have ArcArc in its genotypical constitution. The F2 confirmed this.

Aside from the types virgata and virgarcus, the second generation segregated for a series of types that evidence transitions from maximus to minimus in their spreading of color. Thus it was not possible to clearly separate the four types maximus, major, minor, and minimus, as well as their heterozygotes, according to their genotypical constitution. Therefore, the following overview of the resulting types should only be regarded as an orientation, and the resulting numbers cannot be used in a numeric calculation of segregation ratios. The following results were found:

Partial Color Type

No. of


virgarcus, heterozygote in arcus




maximus, with less color spreading




major, with less color spreading




minor, with less color spreading




minimus, with less color spreading






While the preceding numbers may be completely unsuited to an exact analysis of the observed segregation, they do allow a fairly safe conclusion with regard to the genes involved. In the previous cross, number 174, the gene diff was determined, which in its recessive form transformed the virgata into the major type and the virgarcus into the maximus type. The same types were also found in the present cross. Thus, there is a segregation in the two genes Bip and Diff. However, two further types of partial color, minor and minimus, were observed. The hereditary consistency of the types had already been determined (LAMPRECHT, 1934); i.e., they appear in homozygote form. In the present cross, then, there most be another gene segregating that is responsible for the formation of those two types and that was clearly introduced by the one partially colored parent, the minor type. Since the two partial color types with the greatest spreading of color on the testa, the minor and the minimus types, occur in smaller numbers than the maximus and major types, and since the minimus type, which shows the greatest spreading of color, is also the rarest, it would seem that the effect of the gene in question is similar to that of the gene diff, i.e., it causes greater distribution of color in recessive form. This gene will be labeled with the symbol Exp, derived from expandere = spreading.

A delineation of the different heterozygotes versus the homozygotes in the series of types maximus - major - minor - minimus will only be possible in crosses with segregation in only one sole gene for partial color, and even then perhaps only through quantitative methods.


  1. To increase our knowledge of the inheritance of partly colouring of the seed coat of Phaseolus vulgaris four crosses were studied.
  2. The two types of partly colouring bipunctata and virgata (Fig. 1) differ in the gene Arc. The genotypical constitution of the bipunctata-type is bipbip arcarc, that of the virgata-type is bipbip ArcArc. Heterozygosity in Arc causes the formation of an intermediate type (Fig. 1).
  3. Through yet another gene, Bip, in co-operation with Arc, the following types of partly colouring arise: arcus, BipBip arcarc (Fig. 3), virgarcus, BipBip ArcArc (Fig. 2). Heterozygosity in the gene Bip also causes intermediate types (Fig. 4).
  4. While the two genes Arc and Bip in their dominant state cause the extension of colour on partly coloured types, the two new genes Diff and Exp do so in their recessive state. These two genes Diff and Exp are responsible for the formation of the types: maximus, major, minor and minimus (Figs. 5, 7, 9, 10). Heterozygosity in these genes causes also intermediate types (Figs. 6, 8, 11). In such cases the result will be a series of gradual transitions, as e.g. in cross No. 282 from maximus to minimus.


  1. Lamprecht, H. 1934. Zur Genetik von Phaseolus vulgaris, VIII. Über Farbenverteilung und Vererbung der Teilfarbigkeit der Testa. Hereditas, XIX, 177-222.
  2. Sax, K. and McPhee, H.C. 1923. Color Factors in Bean Hybrids. Journ. of Heredity, XIV, 205-208.
  3. Schreiber, F. 1934. Zur Genetik der weissen Samenfarbe bei Phaseolus vulgaris. Der Züchter, VI, 53-61.
  4. Surface, Fr. M. 1916. A Note on the Inheritance of Eye Pattern in Beans and Its Relation to Type of Vine. Amer. Nat., L, 577-586.
  5. Tschermak, E. v. 1912. Bastardierungsversuche an Levkojen, Erbsen und Bohnen mit Rücksicht auf die Faktorenlehre. Z. f. ind. Abst. u. Vererbgl., VII, 81-234.