W. Bateson, in 1905, described a cross in sweet pea, where a deviation from independent assortment was exhibited. Plants of a sweet pea variety having blue flowers (B) and long pollen (L) were crossed with those of another variety having red flowers (b) and round pollen (l). F1 individuals (BbLl) had blue flowers and long pollen. These were crossed with plants having red flowers and round pollen (bbll). (In this case character for blue colour of flowers is dominam over red colour and long pollen character is dominant over round pollen). Normally if independent assortment takes place, we should expect 1:1:1:1 ratio in a testcross. Instead, 7:1:1:7 ratio was actually obtained, indicating that there was a tendency in dominant alleles to remain together. Similar was the case with recessive alleles. This deviation was, therefore, explained as gametic coupling by Bateson. Similarly, it was observed that when two such dominant alleles or two recessive alleles come from different parents, they tend to remain separate. This was called gametic repulsion.

In Bateson's experiment in repulsion phase, one parent would have blue flowers and round pollen (BBll) and the other would have red flowers and long pollen (bbLL). The results of a testcross in such a repulsion phase were similar to those obtained in coupling phase giving 1:7:7:1 ratio instead of expected 1:1:1:1. Bateson explained the lack of independent assortment in the above experiments by means of a hypothesis known as coupling and repulsion hypothesis. Although coupling and repulsion as explained above were later discovered to be the two aspects of the same phenomenon called linkage, the terms coupling phase and repulsion phase are still considered to be useful terms in scientific literature.

T.H. Morgan, while performing an experiment with Drosophila, in 1910, found that coupling or repulsion was not complete. He proposed that two genes are found in coupling phase or in repulsion phase, because these were present on same chromosome (coupling) or on two different, but homologous chromosomes (repulsion). Such genes are then called linked and the phenomenon is called linkage. He further suggested that the strength of linkage will, be determined by distance between two genes in question. The greater this distance, lower will be the linkage strength. The linkage is broken down due to the phenomenon of crossing over occurring during meiosis. Crossing over will be relatively more frequent, if distance between two genes is more than in a case where the distance between two genes is less. The phenomenon of crossing over involves exchange of chromosome segments (Fig. 10.1).

Fig. 10.1. Exchange of chromosome segments during crossing over. The first figure represents a pair of normal chromosomes, the second figure represents exchange due to double crossing over.

In order to make the phenomenon of linkage and crossing over easily understandable, let us take another hypothetical example, where genes A and B are involved, a and b being their recessive alleles respectively. A cross AB/AB x ab/ab would give rise to a F₁ dihybrid AB/ab. This dihybrid will then be crossed with double recessive parent ab/ab to get the testcross progeny. The formation of gametes from dihybrid (AB/ab) in presence and in absence of crossing over is shown in Figure 10.2. The derivation of test cross progeny is also represented.

Fig. 10.2. A testcross (AB/ab x ab/ab) showing crossover and non-crossover gametes (from AB/ab) and progenies.

Depending upon distance between any two genes which is inversely proportional to strength of linkage, non-crossovers will vary from 50%-100% (100% non-crossovers is a state where no crossing over takes place as in male Drosophila. The crossovers will similarly vary from 0-50% and will never exceed 50%. The easiest way of finding out the proportion of non-crossovers and crossovers is to make a testcross (AB/ab x ab/ab) as shown in Figure 10.2. In such a situation, since phenotypic ratio will depend upon ratio of different kinds of gametes coming from F₁ (AB/ab), the relative proportion of non-crossovers and, crossovers can be easily determined from phenotypic ratio. This is illustrated in the next section using an example from maize.