Carlo Maria Giulini conducts L’Orchestre de Chambre de Hollande in a performance of Mozart’s Symphony No. 35, “Haffner,” on August 27, 1959. I don’t think I’ve seen footage of Giulini this early in his career. He’s 45 here. This video was only recently added to YouTube, and few have watched it. What do you think?

When did it stop being a custom to slow up in second subjects? I miss that.
Yeah, that’s nice, I agree. When did it stop being a custom? Hard for me to say, but I would guess that the historical performance practice guys helped putting a stop to it. btw, if memory serves, Gustavo Dudamel slows down for the second subject in Mozart. Part of what I call his Old School approach.
A more interesting question for me would be, when did it START being a custom? Because, as far as i know, there is no evidence that Wolfgang Amadeus wanted to slow down so considerably in his “second subjects”. And i am glad that our beloved Maestro is no longer alive, because, if he heard that his interpretation of Mozart was “customary”, it probably would have hurt him an awful lot.
As Tim said, those Period Performance Practice People can probably be credited (or, depending on your point of view, blamed) for cleaning this thing up (which in my opinion is one of their more positive accomplishments), but only partly so, because, even before Giulini and long before PPPP time, his older compatriot, a guy named Arturo Toscanini, already preferred much more unified tempos in his Mozart: listen for example to where the “second subject” is played at almost exactly the initial tempo of the movement. And, by the way, Giulini himself, a couple of decades later in his life, moved the two speeds much closer together than in the recording posted by Tim above here.
As for my overall impression of this recording – well, we can certainly see the same musician with same musical sensitivity as the one we knew here in LA from late 1970s and until mid 1980s, but detailed comparison is difficult to make because of the inferior quality of the recorded sound and the less-than-stellar level of the orchestra led by him in this performance. However, it seems to me that he became a wiser a more refined musician by the time he arrived in Southern California in 1978. For example, he realized by then that creating different musical characters within one symphonic movement from the classical period does not require drastic tempo changes.
A purely metronomic Mozart would certainly be very unattractive, but the degree of tempo fluctuations and rhythmic freedom that would be ideal in his music remains a crucial matter that will, i think fortunately for us, never be decided definitively.
The Toscanini recording i was referring to is at
youtube.com/watch?v=t1T2v_nawtk .
I found this fascinating for the quite dramatic tempo changes noted above as well as seeing Giulini’s conducting style in those years and what developed. Even though 1959 was still a time considered “old school”, I don’t recall Walter doing something similar. Even Klemperer’s legendary Magic Flute recording from a few years after this does not, I believe, have such changes in it. I also think, as MarK points out, that Giulini did not (or would not) do it this way by 1970s onward based on recordings I’ve heard. I remember Giulini saying that no matter how many times he did a piece, he would always look at it as if it were the first time. So he probably realized at some time that he did not need to do these tempo changes in order to achieve a satisfying performance.