This is an action to recover damages to plaintiffs’ land resulting from the alleged withdrawal of lateral support. Plaintiffs appeal from a directed verdict in favor of defendants. The question on appeal is whether there was sufficient evidence to raise a submissible issue of the defendants’ strict liability for the removal of lateral support.
The principle of lateral support with which we are here concerned is stated in 4 Restatement of Torts § 817, p. 187 (1939):
“(1) Except as stated in § 818, a person who withdraws the naturally necessary lateral support of land in another’s possession, or support which has been substituted for the naturally necessary support, is liable for a subsidence of such land of the other as was naturally dependent upon the support withdrawn, in the absence of a superseding cause or other reason for relieving him.”
As Comment c explains, “Naturally necessary lateral support is that support which the supported land itself requires and which in its natural condition and in the natural condition of the surrounding land it would require.”
It is undisputed that defendants’ withdrawal of lateral support caused plaintiffs’ land to subside. The only question is whether the alterations in plaintiffs’ *517land were a substantial factor in causing their land to subside. If plaintiffs’ land would have subsided even if the alterations on plaintiffs’ land had not occurred, defendants would be strictly liable for the subsidence.① We turn then to a consideration of the facts.
In 1967 plaintiffs purchased a three-acre tract in Clackamas County. Plaintiffs constructed a dwelling house on the top of a hill located on the tract. In conjunction with this construction, plaintiffs dug a drywell② and subsequently cut out a dog run by removing from the hillside a strip of ground 100 feet long and five feet deep. The cut and the drywell were downhill from the house. The presence of plaintiffs’ house had the effect of diverting a part of the natural distribution of rainfall on the hillside. The eave troughs on plaintiffs’ house drained to the rear of the house. The trough serving the left side of the house drained into the drywell; the trough serving the right side drained the water down the hillside.
In 1971 plaintiffs sold several lots located at the base of the hill below plaintiffs’ residential site to defendants. After completing construction of a house on one lot at the base of the hill in October, 1971, defendants excavated a triangular section of the toe or base of the hill about six feet into the hillside. This excavation left a vertical drop from the natural slope of the hillside of between four and ten feet. Defendants had also removed trees and brush from the hillside in the process of constructing their home.
*518Shortly after defendants completed their excavation at the toe of the hill, the first of a series of landslides on plaintiffs’ land occurred. The slides were progressive in nature, each slide claiming additional ground' as the successive slides moved up the hillside. Defendants constructed a retaining wall in March, 1972, at the site of the original excavation, in the process of which they dug another three or four feet into the hillside. Another slide occurred almost immediately thereafter. Eventually, the slide activity reached the top of the hillside and affected an area 45 feet wide and 60 feet long. No damage was done to plaintiffs’ house and they seek to recover only for the damage to the land itself.
Defendants contend that plaintiffs’ conduct contributed to the subsidence in three different ways: (1) by the draining of water into the drywell; (2) by the excavation made for the dog run, and (3) by altering the pattern of rainfall run-off. There is no evidence that plaintiffs’ house contributed to the subsidence.
Plaintiffs’ evidence concerning the effect of these factors was supplied by Badley Squier, a consulting engineer in soil mechanics. Mr. Squier conducted a series of tests on the hillside to determine the cause of the landslides. As to the drywell, Squier testified that neither the type of materials contained in it nor its water-collecting function had any effect on the weight of the soil at the toe of the hill. lie also reasoned that “if this well was causing the landslide activity the very first slide would have come back and engulfed the dry well.” Thus, Squier concluded that neither the weight nor water-collecting activity of the drywell affected the stability of the hillside in a causal sense.
*519As for the dog run, Squier stated that “the cut itself probably helped [prevent sliding] as far as weight is concerned.” He stated that the extra weight of water accumulated because of the cut was not greater than the weight of the ground removed. Thus, he concluded that the cut was not a factor in causing the land to subside.
As for the altered rainfall run-off pattern, Squier testified that the water falling directly on the slope was 15 times greater than the amount deposited by the downspout and concluded that “the rainfall effect and the natural ground water seepage effects were overwhelming with respect to what amount of water to be let into the landslide area from the downspouts and far outweighed any influences of those sources of water.”
Squier pointed out that, given the soil composition and natural slope of the hillside, the area had been marginally stable originally, but that after the excavation at the toe the stability of the slope was greatly decreased, “indicating a retaining wall had to be provided to prevent the failure of that slope.” This led Squier to conclude: “In my opinion the direct cause has to be the excavation made at the toe of the slope in the development of the Busic property.”
In describing the relative effect of the parties’ respective conduct, Squier testified as follows:
“A First of all I would say that the primary cause of the landslide in my opinion based upon my study is the excavation of the toe. We had a series of slides starting at the toe and eventually getting back up to the top of the hill. At the time that may be the fourth or fifth slide had developed and we are now talking about encompassing the ground *520where the dry well was and encompassing the.rnn where the dog rnn bench was. At that time it was a contributing factor to the problem. But it certainly in my estimation is not the cause for it — it’s not the trigger. In other words that sent the hill in motion.. It’s the excavation of the toe.
“Q So this was only a factor when the slides go to that point?
“A It’s a factor or influence, that is correct.”
It seems clear that Squier’s use of the term “primary cause” was not intended to suggest that it was only one of the factors in the original subsidence of plaintiffs’ land at the toe of the hill and that plaintiffs’ conduct was also a contributing factor. The witness concedes that after the series of slides had advanced to the point where the drywell was located, “at that time it was a contributing factor.” (Emphasis added.)
Although plaintiffs are not entitled to recover for the damages caused at the point where their conduct contributed to the subsidence, they are entitled to recover for all the damage that occurred prior to that time.
Since there was evidence to support plaintiffs’ claim based upon the theory of strict liability, the trial court erred in directing a verdict for defendants.
Reversed and remanded for a new trial.
5 Powell on Real Property § 699, 285-286 (1971).
A drywell is described as a channel filled with gravel and covered with soil.