There are plenty of Flagstad/Melchior broadcasts out there, and most often with Bodanzky conducting, so after a number of listenings we can certainly identify selected characteristics of the performances, and begin to nurture the illusion that we know what they must have been like. And if we have a good deal of broadcast-to-studio-to live listening experience, that lends a measure of credibility to the illusion, particularly with respect to the voices. The gap is still wide, however, especially with broadcasts that present sonic challenges, and especially with the orchestra. Its sonic presence in the house (and therefore a significant portion of our impression of what sort of orchestra, under what sort of leadership, we are hearing) we must to a greater or lesser extent infer, and with that inference comes our own preconceptions, derived from whatever a priori knowledge we may possess, whatever memories we may cherish, whatever opinions we may already have formed. There is ample room for slippage there, but among the things we can determine with a fair assurance of accuracy are the tempi. I’ve seen several references, including one (along with the citation of a corroborative review) in the booklet essay by Jeffery S. McMillan, to Bodanzky’s quick pacing, beginning with the Act 1 Prelude (“relatively fleet,” in McMillan’s words). I wish I knew what he and the others were talking about. Bodanzky’s Prelude is notably, epochally slow. To compare like to like, it comes in at 15:48, a minute and forty-four seconds slower than Fritz Busch’s on the ’36 Colón performance (14:04), which in turn is one second slower than Wilhelm Furtwängler’s of March, 1938 with the Berlin Philharmonic. This last is not quite like-to-like, inasmuch as it is a studio recording, therefore sonically somewhat superior, but potentially influenced by side constraints—though I think that’s not the case here, since the recording was originally spread over four 12-inch 78 rpm sides.^(I) To broaden the view a bit, I also timed the Prelude recorded live at Bayreuth in 1962 under Hans Knappertsbusch, famously no speedster. I was surprised to find that from first note sounded to that first reveille trumpet, it comes in at twelve minutes flat, to which we may add two or three seconds to allow for my wristwatch timing. I was so surprised, in fact, that I grew suspicious of the duration given for the Marston track. So I relistened to that, and over the same stretch registered a 15’44″—given the same allowance, a confirmation of the Marston time.^(II)