While studying the chromosome theory of heredity in 1911, American geneticist Thomas Hunt Morgan occasionally noticed that "linked" traits would separate. Meanwhile, other traits on the same chromosome showed little detectable linkage. To explain his results Morgan proposed a process of crossing over, or recombination. Specifically, he proposed that the two paired chromosomes could "cross over" to exchange information. Morgan also proposed that Mendelian factors (genes) are arranged in a linear series on chromosomes, "similar to pearls on a string." He hypothetized that the interchange of genetic information broke the linkage between genes. The closer two genes were to one another on a chromosome, he theorized, the greater their chance of being inherited together. Conversely, genes located farther away from one another on the same chromosome were more likely to be separated during recombination. Therefore, Morgan correctly proposed that the strength of linkage between two genes depends upon the distance between the genes on the chromosome.
Morgan, "Random segregation versus coupling in Mendelian inheritance," Science 34 (1911) 384.
J. Norman (ed) Morton's Medical Bibliography 5th ed (1991) no. 245.3.