My Italian colleague Roberto Delli Carri shared with me his favorite notes from the Tosi book edited by Luigi Leonesi that I mentioned earlier and I couldn’t resist sharing them here too:

Many have always believed and still believe, that at the end of a register there exists a passaggio and there begins another equal register of perhaps a greater force. No: it is not in this sense that the vocal mechanism responds to all the needs of bel canto. But rather by combining the weaker sound (head) with the strong (chest), in a way to be able to obtain from the two different registers such a sweet connection, whereby the singer, for example, can sing a messa di voce, can pass from one register to the other in an imperceptible manner to the listener, as the voice rises from one register to the other.

Having conquered such a great difficulty, no more passaggi will exist in the registers, but tones that are perfectly united instead.

Though Tosi and Mancini explicitly declare that there are only two registers (the natural [or chest] voice and the other the artificial voice [falsetto/head]), the predominant opinion today however is that there are three registers. This error comes from not knowing today how the vocal mechanism should act in song, much less what means must be used to achieve the union of the registers. Given such a state of things, it follows that from the beginning the natural (chest) voice struggles to surpass the limits assigned to it, and in so doing, the voice disintegrates and breaks in two, and here is the reason that we find another register. The Art lies in knowing how to combine these two registers and if the union of them is not perfect, the voice will be of many registers.

Tosi, Pier Francesco. Opinioni de’cantori antichi e moderni. No. 50. Forni, 1904.