MURRAH, Chief Judge.

In this diversity suit from Kansas, Mrs. Derr, a regular customer at a Safeway Store, had just lifted a carton of soft drinks from a stack and had turned to put the carton in her grocery cart when a carton of Pepsi Cola fell from an adjacent stack, injuring her foot. She brought suit against the Pepsi Cola Bottling Company and the Safeway Store, claiming that Pepsi Cola negligently used spring roll shelving 1 to stack cartons; that Safeway negligently allowed their use in its store; and that the snap back of the shelving on which her carton rested struck the “improperly stacked” adjacent carton, causing it to fall. The jury returned a verdict in favor of Pepsi Cola but against Safeway. The trial court sustained Safeway’s motion for a judgment n. o. v. for insufficiency of the evidence. On appeal, Mrs. Derr asserts the trial court erred in granting the judgment n. o. v. and, alternatively, if the judgment n. o.v. is sustained on the record facts, she is entitled to a new trial on the basis of improperly excluded expert testimony. We know, of course, a judgment n. o. v. is justified only if the evidence is so overwhelmingly preponderant in favor of the movant as to admit of no other reasonable conclusion. See Gulf Insurance Company, Inc. v. Kolob Corporation, Inc., 404 F.2d 115 (10 Cir. 1968); Readnour v. Commercial Standard Ins. Co., 253 F.2d 907 (10th Cir.). Cf. C. H. Codding & Sons v. Armour and Company, 404 F.2d 1 (10th Cir. 1968).

Mrs. Derr does not rely upon res ips a loquitur to prove negligence; instead she contends that the very use of spring roll shelving created a dangerous condition. Indeed, the trial court told the jury without objection that Mrs. Derr sought to recover damages on the theory that “the defendants were negligent in that the Pepsi Cola Company stacked cartons of bottles in the store by the use of a plastic material known as ‘Mylar’, that Safeway was negligent in maintaining such stacks in its store and failing to warn the plaintiff of the danger that cartons would fall from the stack; and that both defendants should have foreseen that injury to Safeway’s customers would likely result from the manner in which the cartons were stacked.”

Kansas follows the generally accepted rule which imposes upon storekeepers the duty to provide their customers a reasonably safe place to shop and to warn them of any unreasonable risk of harm of which the storekeeper should be aware and which is not readily apparent to the customer. 2 See Little v. Butner, 186 *637 Kan. 75, 348 P.2d 1022, citing Restatement of the Law, Torts, §§ 342, 343. See also 20 A.L.R.2d 95-111 for cases on stacking and exhibiting merchandise. Thus, our crucial negligence question is whether the evidence was sufficient to show that the use of spring roll shelving created such an unreasonable risk of harm as to impose upon Safeway the duty to take necessary precaution to either avoid the risk or to adequately warn its customers of the lurking danger.

The evidence of negligence showed without dispute that the adjacent carton did fall, that in the ordinary course of a shopping day customers in the store pick up cartons and replace them, and that a soft drink carton falls about once a week. In this respect the theory and facts of our case are somewhat akin to Gonzales v. Shoprite Foods, Inc., 69 N.M. 95, 364 P.2d 352 (1961), where the New Mexico court held the fact that several articles fell when a customer lifted one from the top of a stack did not give rise to a rebuttable inference that the customary manner of stacking in and of itself constituted negligence. See also Graham v. Fed-X, Inc., 384 S.W. 2d 785 (Tex.1964). But, in Jiffy Markets, Inc. v. Yogel, 340 F.2d 495 (8th Cir. 1965), the Eighth Circuit, applying Kansas law, held the evidence of negligence sufficient where a customer at a grocery store was injured by walking through a large glass panel which, formed part of the front of the store. The court did not think the jury was “required to speculate as to the dangerous and unsafe condition created by the glass front. There was evidence to that effect. A former employee of defendant testified that during a period of eight months he had observed four or five persons come in contact with the glass front and ‘bounce off’. A safety engineer testified it was a hazardous arrangement, and detailed methods that could have been employed to correct the visibility of the glass.” And cf. Blohm v. Cardwell Mfg. Co. 380 F.2d 341 (10th Cir. 1967).

And so it is with the evidence in our case. A safety engineer called by Mrs. Derr as an expert witness testified he had experience with safety requirements in the stacking and exhibiting of merchandise and that the factors to be considered include the stability, weight, and size of the unit to be stacked. He had examined the general stacking arrangement on the spring roll shelving at the, Safeway Store about a year after the accident; had duplicated the same general arrangement at another location based upon his examination; and had conducted experiments “to test their stability by jostling and by lifting a carton off the top and replacing it askew as it many times is in situations of this sort.” This quoted testimony was stricken as “speculative”. Despite much wrangling with the trial court this expert was not permitted to testify as to his conclusions about the safety features of spring roll shelving as used in the stacking of Pepsi Cola at the time of the accident because, as the trial court ruled, his experimental hypotheses were not shown to be sufficiently similar to the conditions at the time of the accident. 3 See Annotations 85 A.L.R. 479. But after extended colloquy and after much probing and changing of positions, 4 the expert was *638 finally permitted over Safeway’s objection to express his opinion regarding the safety features of spring roll shelving in general, that it was a “non-rigid, unsupportive material not capable of stabilizing merchandise piles.” 5 If as indeed we think, the jury should be permitted to believe that the spring roll shelving was an unsupportive material not capable of stabilizing merchandise, they could also believe, as they apparently did, that its very use created a potentially dangerous condition. And considering that a carton fell from the stacks about once a week, for whatever possible reason, the jury could also believe that Safeway should have been aware of the dangerous condition with consequent duty to warn. We are convinced that this evidence made out a case of negligence for the jury, but the existence of a dangerous condition and failure to warn prove only negligence, not that the negligence proximately caused the adjacent carton to fall upon Mrs. Derr’s foot.

To prove proximate cause Mrs. Derr contends the evidence was sufficient to justify an inference that the snap back of the shelving upon which her carton rested struck the adjacent carton with sufficient force to cause it to fall. Implicit in and essential to this theory is the hypothesis that the adjacent carton which fell was “improperly stacked”, i. e., “askew”. 6 From its rulings on the evidence and the judgment n. o. v. it is manifest that the trial court felt there was no evidence to justify the hypothesis that the adjacent carton was askew. 7 And absent such evidence or permissible inference, an experiment based upon this hypothesis was inadmissible to prove either negligence or proximate cause.

Kansas adheres to the general rule that for evidence based upon an experiment to be admissible, the conditions under which the experiment was carried out need not be identical with those existing at the time of the accident. It is sufficient if there is a substantial similarity between these conditions. Pool v. Day, 143 Kan. 226, 53 P.2d 912; Wegley v. Funk, 201 Kan. 719, 443 P.2d 323; 29 Am.Jur.2d, Evidence § 824. Cf. Millers’ National In *639 surance Co. v. Wichita Flour Mills, 10 Cir., 257 F.2d 93. The underlying reason for the rule is that to be admissible testimony of this kind must tend to aid and assist the jury in its consideration of the evidence and the parties’ theories of the case, and. not tend to confuse. Blohm v. Cardwell Mfg. Co., 380 F.2d 341 (10 Cir.). The trial court is, of course, the first and best judge whether conditions of the experiment are sufficiently similar and enlightening to render the testimony based thereon admissible. And we must not disturb the court’s ruling on these critical issues unless we are convinced it is clearly wrong. See Millers’ National Insurance Co., supra.

This brings us to consideration of the legally permissible inferences to be drawn from basic facts. While Mrs. Derr did not notice any carton being askew, it was shown that she was the only person near the soft drinks; that she lifted the carton straight up and did not touch any other carton; that customers removed and replaced cartons while shopping; and that the adjacent carton fell a “split second” after the shelving snapped back. With great respect for the consistency of the trial court’s evidentiary rulings, we are constrained to believe that when the evidence is construed most favorably to Mrs. Derr, the facts and circumstances of this case are entirely sufficient to permit the inference that the adjacent carton was askew at the time of the accident. Indulging this inference the conditions of the experiment become sufficiently similar to the conditions at the time of the accident to be relevant. The integrity of the hypothesis and the conclusions based thereon are, to be sure, vulnerable to attack on cross-examination and rebuttal and their ultimate weight and value is of course for the jury. 8

While the inference that the adjacent carton was askew is permissible, Mrs. Derr must yet show to the satisfaction of the jury that the “snap back” of the spring roll shelving had sufficient force to knock off an eighteen pound, eight bottle carton of Pepsi-Colas. Without such evidence, a crucial link in the chain of causation is missing, and on the record facts the judgment n. o. v. is appropriate even though, as the trial court instructed, proximate cause is usually for the jury. Responsively to this issue, the safety expert testified there was “sufficient force in the snap back to, in fact, knock a carton of Pepsis off under certain circumstances.” But this testimony was stricken as speculative, apparently because of the use of the phrase “under certain circumstances.” This phrase is, indeed, too vague on which to predicate an admissible judgment on whether the snap back caused the carton to fall. But when the phrase “under certain circumstances” is shown to refer to the inferable fact that the adjacent carton was askew, the conditions become sufficiently similar to make the experimental testimony admissible; supply the missing link; and take the case to the jury on proximate cause.

Despite Mrs. Derr’s failure to file a motion for a new trial within ten days after entry of the judgment n. o. v. according to F.R.Civ.P. 50(c) (2), this, court is empowered to grant a new trial if the jury verdict is not reinstated. Neely v. Martin K. Eby Const. Co., 386 U.S. 317, 328, 87 S.Ct. 1072, 18 L.Ed.2d *640 75, citing The Advisory Committee’s Note to Rule 50(c) (2), 31 F.R.D. 646.

The case is remanded for a new trial in accordance with the views herein expressed.

. Spring roll shelving, sometimes referred to as mylar rollers, are paper thin sheets of plastie, several ounces in weight, which are laid between layers of soft drink cartons to stabilize the stacks. When a carton is picked up from a stack, the shelving rolls up rapidly or “snaps back” against the carton directly behind it in the same row of cartons.

. The jury was instructed without objection that although Safeway was not an insurer of Mrs. Derr’s safety, it did owe her “ * * * the duty of exercising *637 ordinary care to keep the portion of the premises ordinarily used by customers in a reasonably safe condition, and it had a duty to warn customers of dangerous conditions which wore known, or which reasonably should have been known to it but not to the customers.”

. Explaining his ruling, the trial court states:

“I’m not going to permit this man to testify to conclusions based on experiments he made placing cartons in different positions where there is no evidence here that there was anything of that sort and where the evidence is to the contrary. This is pure — it’s not an opinion based on anything in evidence or which can be inferred from the evidence.”

. At one point in his proffer of proof, plaintiff’s counsel stated: “ * * * we would like to make an offer of proof in this case that this witness is an expert in the engineering field and the stacking of merchandise, and also as an expert, *638 he’s entitled to make an opinion as to whether mylar rollers are safe or unsafe in the stacking ■ of Pepsi-Cola, and the purpose of introduction of testimony on the investigation is to lead up to his final conclusion as to his investigation and his experiments relating to stacking of Pepsi-Cola on mylar rollers.

“Now, we believe that he is not going to in any way through his testimony relate conclusions to the specific act as it occurred up there, but just to mylar rollers, and the fact that they are dangerous, as such, to use, dangerous by themselves, and they are unsafe.”

. This answer was in response to the question, “What did you find in your investigation relating to mylar rollers?” The trial court properly admitted'' this testimony without a hypothesis for a recent Kansas statute provides: “[Q]ues-tions calling for the opinion of an expert witness need not be hypothetical in form unless the judge in his discretion so requires, but the witness may state his opinion and reasons therefor without first specifying data on which it is based as an hypothesis or otherwise; but upon cross-examination he may be required to specify such data.” K.S.A. 60-458.

. As a part of her case on negligence and proximate cause, the plaintiff proffered: “But the witness is qualified as an expert to show that if you are going to use this type of stacking device, that it must be carefully stacked and one must be careful to get those Pepsis directly one on top of the other and that if some are slightly askew and a mylar roller does zip back and touch a carton next to it, that the probabilities are that the carton is going to do just what the carton in this case did; it’s going to fall * * * She glanced at it to know where it was to pick it up and picked it up and in placing it in the carrier, the second one fell down, and she says she didn’t knock it down. The only other thing is that it must have be'en, askew sufficiently to be pulled off balance by the snapping of the mylar roller.”

. As we have seen, the trial court ruled that the conditions of the experiment were not shown to be sufficiently similar to the conditions existing at the time of the accident to render it admissible. See footnote 3.

. Indeed, the trial court instructed the jury that if it finds

“ * * * a state of facts materially different from that assumed in the hypothetical questions, then, of course, the jury should give no weight at all to the conclusions or opinions of such expert witnesses based upon such assumed facts.” And the jury was further told that they “ * * * should consider each expert opinion received in evidence in this case and give it such weight as you think it deserves ; and you may reject it entirely if you conclude the reasons given in support of the opinion are unsound.” These instructions are clearly applicable to the proffered hut excluded expert testimony, as indeed were the instructions on circumstantial evidence.