The kind of thing that all do-it-yourselfers assume will not happen to them happened to George Kirstein. He was seriously burned while using chemical products to remove linoleum at his home in Westmont, Illinois. Thereafter, he and his wife brought this products liability suit, seeking damages for his injuries, in an Illinois state court. The Parks Corporation and W.M. Barr & Company removed the case to federal court, where the district judge granted summary judgment, dismissing both defendants, and the Kirsteins appeal only as to Parks.

When Kirstein removed the linoleum from his kitchen, foyer, hallway, and bathroom, he discovered that he also had to remove adhesive left underneath the linoleum on the floor. He purchased an adhesive remover manufactured by Parks and heeded the warnings on the container--he provided adequate ventilation and kept the product away from heat, sparks, and flame in order to eliminate the risk that toxic gases would be produced. He used the product without incident, but unfortunately the adhesive remover left a residue. Checking the label on the product, he found that Parks recommended the use of its lacquer thinner to clean up the adhesive remover residue. He bought a lacquer thinner manufactured by Barr, a product not different from the Parks product in any significant way. He heeded the even stronger warnings on the lacquer thinner label, warnings about the danger of flash fires or explosions. He opened doors, turned down the hot water heater so it would not ignite, and set his air conditioner at 79 degrees.

He used the product, proceeding from the kitchen, through the foyer, into the bathroom, and down the hallway. When he was in the hallway, he closed the door to the outside because it blocked an area he needed to clean. Soon after he closed the door, an explosion occurred. It was later determined that the explosion originated in a furnace return grill located in the foyer adjacent to a door leading to a garage. The ignition source was the electric motor for the air conditioning unit. As the fire spread over the floor, Kirstein started to slip. His shoes were burned off and his feet were incinerated. He was severely burned over 20 percent of his body, with the most extensive burns on his hands and feet.

The Kirsteins sued the manufacturers of the two products Mr. Kirstein was using as well as the retailers from whom he bought the products. The retailers were dismissed by agreement. The suit proceeded--now in federal court--against the manufacturers, both of whom moved for summary judgment. The motions were granted in one of the orders we now review. The others are an order denying a request to present a second expert opinion and an order denying reconsideration of the summary judgment decision.

The Kirsteins have a difficult case to make on the merits. The lacquer thinner is almost certainly what caused the explosion, but its container contains warnings which comply with the Federal Hazardous Substances Act, 15 U.S.C. § 1261, and this preempts the claim against Barr. That's why, it seems, the suit against Barr is not being pursued on appeal. What the Kirsteins claim is that the Parks adhesive remover, while perhaps safe in itself, includes an instruction which renders the product unsafe. That instruction is that for cleanup of the residue of adhesive remover, one should use lacquer thinner. No warning on the adhesive remover container states that use of lacquer thinner is hazardous.

In what has been a rather protean theory, the Kirsteins contend that the combination of the products caused the severity of Kirstein's burns. We will have more to say about the changing nature of the theory later, but for now we need only say that the case was heavily dependent on the opinion of the Kirsteins' expert, Dr. Gary Nelson, regarding the combination of the products. Both defendants moved for summary judgment, in which the admissibility of Dr. Nelson's opinion became a significant issue. Ultimately, his opinion testimony was excluded under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993), and its exclusion became the underpinning of the decision granting summary judgment. Just prior to the ruling on summary judgment, the Kirsteins requested leave to present an opinion from a second expert. That request was denied as was a later motion to reconsider.

We review summary judgment de novo. Porter v. Whitehall Laboratories, Inc., 9 F.3d 607 (7th Cir.1993). However, the Daubert issue is reviewed for an abuse of discretion, General Electric v. Joiner, 522 U.S. 136, 118 S. Ct. 512, 139 L. Ed. 2d 508 (1997).

The district judge carefully analyzed the Daubert issue by employing the two-step inquiry for evaluating expert testimony under Rule 702. First, she looked to see if the expert's testimony is reliable; and, if so, whether it would assist the trier of fact. Cummins v. Lyle Industries, 93 F.3d 362 (7th Cir.1996). She acknowledged that Dr. Nelson had respectable credentials, but she nevertheless concluded that because he was not a chemist (and, in her view, the combination of these products required knowledge of chemistry) and did no testing, his opinion failed the first part of the Daubert test. Based on Rosen v. Ciba-Geigy Corp., 78 F.3d 316 (7th Cir.1996), her conclusion was that she was confronted with "unscientific speculation offered by a genuine scientist."

As a preliminary matter, the Kirsteins claim they were entitled to a hearing on the admissibility of the expert opinion pursuant to In Re Paoli R.R. Yard PCB Litigation, 916 F.2d 829 (3rd Cir.1990). We are convinced, however, that the district court had a sufficient basis for her decision without holding a hearing. We have not required that the Daubert inquiry take any specific form and have, in fact, upheld a judge's sua sponte consideration of the admissibility of expert testimony. O'Conner v. Commonwealth Edison Co., 13 F.3d 1090 (7th Cir.1994).

Furthermore, we see no abuse of discretion in the exclusion of the expert. It is true that Dr. Nelson has impressive credentials. He is a safety engineer. His educational background is in industrial and product safety. He has a Ph.D. from Texas A&M University. He was vice president of the Texas Safety Association and is a member of the National Fire Protection Association. But the fact is that he did no testing on these products, either alone or in combination. Neither did he provide studies which employed such testing. In short, Dr. Nelson offered only speculation. And we have sanctioned the exclusion of speculation offered by persons with credentials as impressive as those of Dr. Nelson. In Rosen, for instance, the expert whose exclusion we upheld was a "distinguished cardiologist and department head at the University of Chicago." 78 F.3d at 318.

The fact that Dr. Nelson did no testing may account as well for what we have referred to as the protean nature of the theory of what happened here. The amended complaint says Kirstein "read and carefully followed the labels on the products to the best of his ability when suddenly the products and the combination of them exploded, causing serious injury." But the initial report of Dr. Nelson, which dealt primarily with the inadequacy of the warnings on the products, did not clearly mention any increased danger from the combination of the products. The closest he came is a statement that "Barr and Parks apparently failed to reasonably and properly foresee the circumstances (or readily conceivable similar circumstances) that resulted in Mr. Kirstein's injury." We find the theory of an increased danger from the combination of the products in Dr. Nelson's deposition. At the time of the deposition the theory seemed to be that lacquer thinner (methyl ethyl ketone, or MEK) by itself would flash but not continue to burn. But if it is mixed with something else, such as cloth or adhesive remover, it will continue to burn:

Answer by Dr. Nelson: If we have vapor in this room, and we walk in, and there's a flash, we're going to get our eyebrows singed, our hair singed, and we're going to get some burns. If I take MEK and I mix it with something, and I spread this stuff on your skin, and that flash ignites the MEK within that substance, and it's continually giving off vapor from it, it will continue to burn. If it's in your clothing, clothing will continue to burn. If it's in your shoes and on your shoes and on your feet, it will continue to burn, and if it's mixed within a gooey substance, that gooey substance will continue to burn. We're no longer now talking about just a flash and the fire is gone. We're talking about a continually burning substance, and if you look at the [photographs], this is exactly what happened.

In the initial brief to this court the theory as to the combination of the products is stated as follows:

From a danger point of view, mixing the two products creates an enhanced danger.... Photographs of the burn disclose that the mixture stuck to the skin and ignited, and continued to burn.... If you pour MEK on adhesive remover and mix it up and scrape it off, and it gets on your hands and clothing, then you are going to have MEK mixed in with the adhesive remover.... If you pour MEK onto a permeable surface, the MEK will continue to burn after the initial flash.

But in the reply brief the plaintiffs explain their theory this way:

At no time did the plaintiff argue that the lacquer thinner became more flammable or changed its propensities. Plaintiff had always argued instead that the adhesive remover became more flammable and created a dangerous condition by using the lacquer thinner.

It seems to us that experimentation on the products in combination might have helped to pin down exactly what it is that happened--did the adhesive remover itself become more flammable when it was mixed with the lacquer thinner? was it the adhesive remover that burned? or did the lacquer thinner burn longer and more dangerously when it was mixed with adhesive remover? What we have is indeed "unscientific speculation."

When they were confronted with the possible loss of their expert, or at least when it became clear that the admissibility of his testimony was under serious attack, the Kirsteins attempted to submit the testimony of a second expert, who had, in fact, performed tests on the products. The district judge refused to allow the submission of his report. The plaintiffs contend that this decision also was an abuse of discretion.

The Kirsteins say they were not given a chance to respond to Parks' attack on Dr. Nelson because the attack came in reply briefs on the motion for summary judgment. We disagree. The issue of the competence of the expert was clearly raised in the original briefs. But it is also true that the reply briefs contained a more concerted attack on the expert than did the original briefs. The way the issue was presented comes close to sandbagging. But on the facts of this case, we see no abuse of discretion in the way the judge resolved the issue.

First of all, the fact that Dr. Nelson's theory, whatever it was, should have been submitted to testing seems elementary. It would seem, in fact, that testing might have pinned down which product became more dangerous by being mixed with the other--which product was burning. Secondly, it is not so simple a matter to allow the second report. Discovery was closed; all deadlines were final. The report of the second expert is not self-explanatory; it raises more questions than it answers. If the report were allowed, Parks would no doubt want to depose the second expert and perhaps obtain more expert testimony of its own. The admission of the report, in other words, would cause significant delay, or so the district judge could reasonably believe.

The report we are talking about is from Shelby Parchman, who performed tests apparently checking for BTU's on six different samples: 100% adhesive remover; 5% thinner/95% remover; 25% thinner/75% remover; 50% each; 75% thinner/25% remover; and 95% thinner/5% remover. Our view of the unexplained results is that the BTU's (that is, the measurement of heat capacity) increased as more lacquer thinner was added to the mixture. What, if anything, does that tell us about the cause of Kirstein's burns? Furthermore, the product tested by itself is the adhesive remover; there is no test of 100% lacquer thinner. Would the lacquer thinner by itself have greater or less heat capacity than a combination of the products? In short, would the Parchman report be--or, in fact, has it been--the foundation for yet another theory as to just how the products become more dangerous when used together? The district judge, we think, was well within her discretion not to allow the late addition of another expert to the case. And she was also well within her discretion not to reconsider her rulings.

The Kirsteins put forth a number of other reasons why they should survive summary judgment regardless of whether they have an expert. We reject those reasons. No matter which form it takes, their claim against Parks depends on the combination of the Parks' adhesive remover with lacquer thinner. If there is no evidence of how the products operate in combination, there can be no claim. There is no claim that it was the adhesive remover which exploded; there is no claim that the adhesive remover is more dangerous, by itself, than the ordinary consumer would contemplate. There is no competent evidence as to what, if any, contribution the adhesive remover made to the explosion and the resulting burns. Because Barr has been dismissed, there can be no claim based solely on a lacquer thinner explosion. Furthermore, the lacquer thinner is, in fact, explosive; this is what the warnings on the container are about. And because there is no evidence that what happened would not have happened in the absence of adhesive remover, Parks cannot be held responsible under a res ipsa loquitur theory. For all of these reasons, the judgment of the district court is

AFFIRMED.