However, the tendency of most forms was towards the triple division typified by the aria, with a clearly defined first section, a second section of contrasting and uncertain character, and a third section which, being a restatement of the first, reestablished the tonic key and gave to the piece as a whole a positive order. In the binary forms of Froberger and Chambonnières there is the harmonic embryo of a distinct middle section; namely, the few modulations through which the music passes on its way from dominant back to tonic in the second section. It can be easily understood that composers would make the most of this chance for modulation as they became more and more aware of the beauty of harmony; likewise, that the bolder their harmonic ventures in these measures, the greater was their need to emphasize the final re-establishment of the tonic key. Ultimately a distinct triple division was inevitable, with an opening section modulating from tonic to dominant, a second section of contrasting keys and few modulations; finally a restatement of the first section, as in the aria, but necessarily somewhat changed so that the whole section might be in the tonic key. Such is the harmonic foundation of the triplex form. Such a form makes its appearance in music very shortly after the beginning of the eighteenth century. It seems akin now to the aria, now to the binary form. One may suspect the latter relationship if the first section is repeated, and the second and third sections (as one) likewise. These repetitions are obviously inherited from the binary form. On the other hand, if these sections are not thus repeated, the piece resembles more nearly the aria. Take, for example, an adagio from the second sonata in a set of twelve published by Padre Martini in 1742, written probably many years earlier. These sonatas were republished by Madame Farrenc in the third volume of her _Trésor des pianistes_. The adagio in question is clearly in three sections very like an aria, with the difference that the first section ends in the dominant (in the eighth measure), and the last is consequently changed from the first so that it may end in the tonic. There are no repeats. Far more remarkable is a sonata in C major by D. Scarlatti. It is the eleventh in the Breitkopf and Härtel collection of twenty to which reference has already been made. Here we find a first section modulating from tonic to dominant. This is repeated. Then follows the second section, full of free modulations, and this section comes to a very obvious half-close. The last section very nearly repeats the first, except for the necessary changes in harmony so that all may be in the tonic key. Scarlatti nowhere else wrote in this form so clearly. Did he merely chance upon it? The wide crossing of the hands marks an early stage in his composing, yet the form is clearly triplex and astonishingly orthodox.