CONLEY BYRD, Justice.

Appellant, Continental Geophysical Company, appeals from judgments entered in five consolidated cases brought by appellee Jeff Adair, Pleas Garner, R. H. Swint, Johnny McAnally, and Homer Wilkins. The allegation is that appellant's seismograph operations partially destroyed the water supply to appellees' wells.

For reversal, appellant sets out six separate points, but since we hold that the proof fails to show a causal relationship between appellant's seismograph operations and the failure of the water supply in appellees' wells, we deal only with that issue.

The record facts show that some time prior to March 13, 1964, appellant drilled ten holes in the vicinity of the Sugar Grove and Dry Creek areas of south Logan County, at a depth of 100 feet, except for one hole 95 feet deep. In each of seven of these holes appellant placed 200 pounds of dynamite, which took up 40 feet of the hole. The remaining 60 feet was filled with gravel. The charges packed in the holes were set off by appellant on March 13, 14 and 15, 1964. The closest test hole to any of appellees' wells was 1,600 feet; the farthest was 6,300 feet. It was also undisputed that in 1963 Logan County was declared a drouth area, the recorded rainfall being 26.41 inches. (Average rainfall from 1945 to 1965 was 43.82 inches.) Wells other than those here involved also went dry and lawsuits are pending against appellant with respect to some of them.

The depth of most of appellees' wells ranged from 28 to 50 feet. Mr. Swint's well, a drilled well, was 90 feet deep.

Mr. Swint testified that his well was drilled in 1964, before appellant's operations; that it was capable of pumping 750 gallons of water per hour and was supplying, in addition to his home, two chicken houses with a capacity of 12,500 brooders each; that his well quit some time in the spring of 1964; that when it got to where it could not furnish water for one of the chicken houses be got a driller to set up over it and lower the depth and did get a little more water; and that he did not know exactly how long after appellant's crews left before he noticed the water was getting low, but it was not too awful long.

Appellee Homer Wilkins testified that he had owned his place for 12 or 15 years; that he had never had any trouble with his well before; and that he first noticed the lack of water in his well during the summer of 1964.

Appellee Pleas Garner testified that his Bell had been in existence to his knowledge for more than 40 years and had not gone dry, and that he first noticed his problem some time in the summer of 1964. He at first said it was probably in October, but later testified that it happened just after the time of the exploration of the defendant company. Ne did not know whether it was 30 or 60 days or what. Ne knew he had run out of water about the time the defendant put dynamite shots off and had been out ever since. He testified that if the (Dynamite had not caused the damage it was funny because that well had been there some 40 years and had not gone dry. He was willing to say that he felt that the dynamite was what did it because it never happened before to his knowledge and he had been around there all his life. He felt the shots were what caused it. He could not say he knew the dynamite caused it but he had that feeling that that was what happened to turn the water off. He was going to say that. He guessed that God Almighty was the only one that really knew what caused it but he thought he knew.

Appellee McAnally testified that he had owned his place since 1946 and the well had never gone dry before, and that it had started going dry in the summer of 1964.

Mrs. Adair testified that she and her husband owned places at both Sugar Grove and Dry Creek; that they had had plenty of water up until appellant's seismograph operations; and that their wells had gone dry in mid-July, 1964. Her testimony was that she knew they set off some dynamite down there; that she heard and felt the blast and could hear the windows shake; and that it was in 1964 some time and right after they did it that the wells all went dry and they went to hauling water

The only testimony about the geology of the area was given by appellant's witness, Dr. Leslie Mack. He testified that all the wells in the town of Sugar Grove were in the alluvial or unconsolidated settlements; that the water supplying these wells did not move in streams as such, but that it moved as a body en masse, as a saturated mass. He stated that you could dig a well almost any place in the Sugar Grove area and find water at the same level — that it is all saturated and there are no underground streams in the area. He did point out that in areas of limestone formation, such as in north Arkansas, you can have a cavernous limestone where water actually flows through fissures and cracks in the rock, but that so far as Sugar drove is Concerned there are no such streams.

Under the facts, therefore, it appears that in 1963 there was a severe drouth; that on March 13, 14 and 15, 1964, appellant detonated its seismograph test holes; and that some time in the spring or midsummer appellees' water wells began to fail. We think the proof is insufficient to show a causal relationship between the detonation of the test holes and the failure of appellees' wells.

In Western, Geophysical Co. v. Mason, 240 Ark. 767, 402 S.W.2d 657 (1966), we had before us a case of damage to a well by seismograph operations similar to those here. But there the proof showed that the well had been damaged once before by similar explosions, and that soon after the explosions the water in the well turned red and muddy and became unfit for use. Here appellees lack the previous experience found in the Western Geophysical case, and the further element of the muddying of the water.

O'Brien v. Primm, 243 Ark. 186, 419 S.W.2d 323 (1967), involved damage to a water well following a sand fracting operation on an oil well. The proof showed that acid had been placed in the oil well in connection With the sand fract job and that shortly after the 3,500-pound pressure had been applied to the oil well the acid had appeared in Primm's water well. The record here shows no proof of the transfer of any substance from appellant's test holes to appellees' wells.

When we consider the distance from appellant's test hole explosions to appellees' wells, and the time lag between the date of the explosions and that failure of the water supplies, we can find no evidence to show that the test hole detonations were a cause which, in a natural and continuous sequence, produced the damages to the wells and without which the damages would not have occurred. Therefore, we hold that the trial court should have directed a verdict in favor of defendants at the conclusion of plaintiffs' case.

However, our ruling herein does not necessarily require that the case be dismissed. In a similar situation in St. Louis S.W. Ry. Co. v. Clemons, 242 Ark. 707, 415 S.W.2d 332 (1967), we said:

"We come now to the question of whether this case should be dismissed or remanded. This court has long adhered to the rule so well reiterated in Fidelity Mutual Life Insurance Co. v. Beck, 84 Ark. 57, 104 S.W. 533 and 1102 (1907). The general rule is to remand common law cases for new trial. Only exceptional reasons justify a dismissal. One of the exceptions is an affirmative showing that there can be no recovery. Pennington v. Underwood, 56 Ark. 53, 19 S.W. 108 (1892). There it was said that when a trial record discloses `a simple failure of proof, justice would demand that we remand the cause and allow plaintiff an opportunity to supply the defect' To the same effect, see Hinton v. Bryant, 232 Ark. 688, 339 S.W.2d 621 (1960)."

Counsel for appellant readily advises this court that lawsuits involving other wells are still pending. It is not impossible that the defects in the proof could be supplied on retrial. We need not speculate on the nature of the proof. But should we dismiss plaintiffs' causes of action here and should other plaintiffs furnish the needed proof, the practical result would be that some people in the community would recover for their wells while the plaintiffs here involved would be denied. This seeming inconsistency in the law would be contrary to the philosophy that justice should not only be fair but appear fair.

Reversed and remanded.

FOGLEMAN, J., concurs.

JOHN A. FOGLEMAN, Justice, concurring. I concur in the majority Opinion and its disposition of the case. My concurrence in the remand of the case is not for the reasons stated in the majority opinion, however. In view of my dissent from the remand in St. Louis Southwestern Ry Co. v. Clemons, 242 Ark. 707, 415 S.W.2d 332, I feel that I should express my reasons for this concurrence.

There is direct precedent for remanding a case in which judgment based on a jury verdict is reversed for insufficiency of the evidence to support it. The touchstone for the departure from the usual disposition in such case is that there be some unusual circumstance to justify the different treatment. It has been held that such unusual circumstance exists when the lack of substantial evidence is due to the failure of the plaintiff to show causation where expert testimony might have supplied the deficiency. In Reynolds Metal Company v. Ball, 217 Ark. 579, 232 S.W.2d 441, the appellee plaintiff sought to recover from the appellant for damage to his farm and pasture land. He contended that the damage was caused by sediment from the alumina extraction plant of the appellant some six or seven miles upstream. The plaintiff failed to prove that the sediment came from the plant and could only testify that he never had any of the sediment until the plant went in. He also failed to prove that the sediment included chemical constituents that were poisonous to vegetation. This court held that the trial court should have directed a verdict, but found that there were special circumstances to justify remanding the cause for a new trial rather than dismissing. The identical factors relied on in that case are the deficiencies in appellees' proof in this case, so I agree that a remand is proper here, as I consider that precedent controlling.