The task allotted to Scarlatti was, therefore, not difficult. He accepted the already established supremacy of strings, but soon realized that three-part writing did not produce even balance of tone. Consequently, he adopted a manner of writing which comprised a division of the violins into firsts and seconds. He added, moreover, an individual part for the violas, and thereby established a canon of phonetics that has been accepted by all erudite composers since his time. It is true that these characteristics of orchestration cannot be said to have originated with him, but his persistent use thereof established a precedent of permanent value. In three-part writing, not only the violoncellos and basses progressed simultaneously in unison or octaves, but also the viola, if present, reënforced the bass in slavish delineation. It is obvious that this practice was the result either of sophism or of indifference and ignorance. And the fact that as late as the eighteenth century no less a composer than Haydn and even Mozart should have continued to frequently employ three-part writing for the strings is certainly a paradox, and tends to prove how circuitous the process of evolution is. However, Haydn and Mozart had such perfect command of florid counterpoint, that no matter what the distribution of string parts might be, the results were invariably effective. Four instead of three notes of a chord being now properly dispersed among the strings,[7] Scarlatti proceeded to enrich his orchestra by a logical employment of wind instruments in pairs. The harpsichord, of course, continued to hold its own, but the Händelian principle of long held notes in the wind against more agile string passages is already to be found in his scores, a principle of which Lulli was also cognizant. But Scarlatti's orchestra was more plastic than Lulli's, and his overtures more purely instrumental. [Footnote 7: Ex. 6.]