Significance of partial melting as a form of melt production

Question

How and when was the significance of partial melting understood to be a significant form of melt production and could explain the variability in melt (and hence igneous rock) composition?

Answer 1

The best place to find an answer to this question is probably Mind over Magma: The Story of Igneous Petrology, by Davis A. Young. In chapter 13, The Mechanism of Differentiation, he explains how the Soret effect, which was viewed as the main mechanism for rock differentiation at the end of the nineteenth century, began to be rejected:

The Soret effect was the most frequently invoked mechanism of differentiation until 1893, when Backstrom issued a paper attacking it. He was soon joined by Harker, Johnston-Lavis, Michel-Levy, Becker, Brogger, Loewinson-Lessing, Schweig, and even Teall.

He then goes on by listing the various other mechanisms proposed by scientists at the time, such as fractional crystallization, and ends with this interesting passage (p. 225, emphasis mine):

Not every petrologist was convinced that differentiation, whatever the proposed mechanism, was the dominant cause of igneous rock diversity. Partial melting, espoused by Becker and Teall, and assimilation, endorsed by Loewinson-Lessing, were also regarded as important petrological processes. Nevertheless, at the beginning of the twentieth century, differentiation was acknowledged to be the leading source of differences among igneous rocks despite a lack of agreement on the nature of the specific physical mechanisms of differentiation and despite the interest in partial melting, assimilation, or even magma mixing.

That would make Becker and Teall the first to recongnize partial melting as an important mechanism explaining the variability in melt composition. However, as Young points out, it was not widely accepted at the time. It would take another few decades. In chapter 16, The Theory of Crystallization-Differentiation, Young describes the contributions of Bowen, and how Holmes disagreed with it, concluding:

Although the details may differ, contemporary petrologists will recognize Holmes' essential idea in current views about the heating and partial melting of upper granitic crust by rising basaltic magmas that have ponded on the underside of that crust. The same arguments were summarized the following year in a paper on the origin of igneous rock (Holmes, 1932).