As it happened, veteran dealers and collectors had anticipated that the Sotheby's evening sale was going to be relatively weak, and were anxious to get it out of the way. According to most observers, the sale was saturated with filler material, and this was obviously a problem. Even minor auction players could see what was going on. With all due respect, much of the art -- the works by Eric Fischl, Barry Flanagan, Red Grooms, Ron Kleeman and Mark di Suvero -- didn't belong in an evening sale.

That's not to say that some works didn't perform admirably. The Andy Warhol red Liz at Sotheby's certainly lived up to expectations, and the Lichtensteins from the Versace collection exceeded them. However, if auctioneer Tobias Meyer and the rest of the Sotheby's team have any hope of catching up with Christie's, they're going to have to re-evaluate their approach. Had they cut 10 to 15 lots from the sale, they would have created a much tighter and more successful auction. In the fall 2005 sale, look for them to use that approach.

At Christie's the next night, May 11, the "real" auction got under way -- and it was spectacular. Christie's had secured the collection of Barbara and Donald Jonas, which was loaded with fresh-to-the-market Abstract Expressionism, and the sale consequently set records -- and restored confidence in the market. One of the more satisfying moments of the auction came when record-breaking prices were paid for two Joseph Cornell boxes. The blue Medici Princess box smashed the previous auction record of $495,000 (set in 1989), selling for $2.5 million. For decades, Cornell had been the most undervalued of all major American artists. No longer is that the case.

Christie's great Edward Hopper experiment -- including an American painting in a post-war and contemporary sale -- turned out to be rather unsatisfying. The work sold for $14 million, under its presale estimate, to the American paintings gallery Berry-Hill, and the audience shrugged. But give Christie's credit for trying to shake things up. The house's other grand experiment, the insertion into the sale of a Diane Arbus photograph, worked out a bit better. However, while the work exceeded its estimate and sold for $228,000, chances are it would have brought the same price in a photography sale.

Presumably, the gang over at Phillips, de Pury and Co., which held its evening sale on May 12, was also pleased with their performance. In emphasizing work created during the two decades, along with spiking the sale with a few blue-chip paintings, Phillips has hit on its own special formula. Of particular note was the record-setting million-dollar price brought by the Richard Prince "Nurse" painting. It's hard to believe that the same painting went for $80,000 only three years ago, when it was originally shown by the Barbara Gladstone Gallery.

So, now what? The biggest problem facing the auctions is their ever-increasing premiums. Currently, when someone bids $200,000 for a picture, they've effectively bid $240,000 because of the 20 percent buyer's premium. The percentage then retreats to 12 percent on anything over $200,000. The houses also charge a premium to the seller -- a sum that can be anywhere between nothing and 10 percent of the sale price.

With this fee structure, the auction houses are essentially playing into the art dealers' hands. A dealer can now tell a potential consignor, "Don't give it to the auctions. I'll sell your work for a 10 percent commission, which allows you to pocket an extra $20,000 (on a $200,000 sale), and you won't have to wait around for several months until the next auction comes along." This is a powerful argument. While testing the auction waters is exciting, collectors increasingly seem to feel that in a market that's found its range, it might be better to work with a dealer. Taking the sure thing, along with the immediacy of the deal, might start to win out over "rolling the dice" at a future auction.

Another addition to the equation was last week's front-page feature in the Wall Street Journal on hedge-fund operators buying art as an investment. The article suggested that these players are partly responsible for the run-up in prices for artists such as Martin Kippenberger and Richard Prince. But the story failed to address a major issue -- what's going to happen when these investors want to cash out? Who's going to be around to bump prices to an even higher level so these speculators can realize a profit?

Still, it looks like the good times will continue at least through the fall. The market may succumb to a self-fulfilling prophecy. That is, if everyone is saying that high prices are here to stay, then prices might in fact remain high -- even in the face of skepticism that they will ever go higher.

Sure, it's true, trees don't grow to the sky. But some trees do grow to be remarkably tall. Just look at the sequoias -- the largest living organism on earth. If we think of de Kooning, Rothko and others art-world giants as sequoias, the art market will be fine. But the minute we fantasize that there's something even bigger out there -- bigger than a sequoia -- then the art market better beware.

For a complete, illustrated preview, see Artnet's signature Upcoming Auctions calendar.