*501Opinion
These are cross-appeals by the insured plaintiffs, Arthur and Mary Strubble, and defendant, United Services Automobile Association, the insurer, from a judgment entered under a California homeowners all-risks of physical loss insurance policy on special verdicts in favor of plaintiffs and against defendant in the total amount of $42,857.79.
Plaintiffs appeal on the sole ground that the total amount awarded them should have been $17,800 greater for reasons which will be explained later. Defendant appeals on the following two grounds: (1) the jury was erroneously instructed regarding the burdens of proof of the parties as to defendant’s liability under the policy; (2) the evidence that plaintiffs’ loss was proximately caused by an earthquake on June 14, 1967, and that the fair market value of plaintiffs’ residence prior to the earthquake was $36,410 is insubstantial.
The Loss
In February 1960 plaintiffs purchased a very large and beautiful Mediterranean-style home with a panoramic view of the ocean for $60,000.1 This home was located on the cliff overlooking Malaga Cove in the City of Palos Verdes Estates. The reason for the comparatively low price was that the cliff on which the home was situated had experienced a slight landslide some months earlier that had largely destroyed a flagstone patio to the north of the house and undermined slightly its northern wing.2
About 10 p.m. on June 14, 1967, an earthquake centered at Whittier, California, some 28 miles from plaintiffs’ home, occurred. It was a rather small one (4.1 at. Whittier and approximately 3.1 at plaintiffs’ home on the Richter Scale) and, although noticeable on the Palos Verdes Peninsula, was of moderate intensity there (perhaps 2.5 on the Mercale Scale).
Nevertheless the next morning plaintiff Arthur Strubble noticed a crack in the brick patio on the landward side of the house. Further investigation revealed that this patio crack was in the ground itself and ran some 30 to 40 feet through various rooms of the house. Two geologists examined the crack *502and the other structural damage to the house and independently concluded that another landslide under the center of the house had started. Officials of the City of Palos Verdes Estates became concerned over the safety of the house and the safety of those on the beach below in the event the entire front of the supporting cliff sloughed away.
Over the next few months the crack in the earth’s surface widened and lengthened appreciably and a distinct downward slope toward the sea developed. The crack and the earth about it became, however, relatively stable. A spring suddenly appeared, though, on the face of the cliff in November 1967. At the urgings of representatives of both the city and defendant, plaintiffs bulldozed off the seaward top of the cliff3 and brought in large timbers and steel beams to shore up the foundation of the house. Plaintiffs’ family moved into the portion of the home away from the cliff and abandoned the seaward part of the house.
Nonetheless, following the winter rains, the landslide under the house accelerated. Plaintiffs demolished the north wing of the house completely in February and March of 1968. Plaintiffs’ very extensive and expensive demolition and foundation shoring efforts proved largely unsuccessful, though, and in August 1968, after the City of Palos Verdes Estates had informally threatened to condemn the house, plaintiffs’ family moved away. Perhaps one-third of the house remained. At the time of the trial of this case in January 1972 a son of the plaintiffs lived in the remainder of the house as a caretaker.
The Expert Testimony Regarding the Cause of the Loss
Because under the policy before us, earthquake (by reason of a special endorsement carrying a substantial additional premium) is an included peril and earth movement otherwise is an expressly excluded peril, the crucial factual issue in this case has always been whether the earthquake of June 14, 1967, was a proximate cause of plaintiffs’ loss. Five expert witnesses testified on this issue. Two defense witnesses, Dr. Leighton, a geology professor and consultant, and Mr. Shields, a consulting engineer, both were of the opinion that the earthquake did not cause the landslide. .The three engineering geologists who testified were divided in their opinions. Mr. Riccio, who examined the home and premises for defendant just after the earthquake, could not tell whether the earthquake had caused the landslide.4 *503Dr. Pipkin, who had done, on his hands and knees, a detailed geological study of the cliff on which plaintiffs’ home stood and who had been studying the cliff’s geology for over 15 years, thought that the earthquake possibly did cause the landslide. Mr. Lockwood, who was on the property numerous times in 1967 and 1968 following the earthquake (commencing within a month thereafter) in an attempt to determine for plaintiffs if the landslide could be arrested and the home saved, was of the opinion that the earthquake was the probable cause of the landslide.
Dr. Leighton conceded that an earthquake might trigger a landslide and that an earthquake “might be the very fine straw that broke the camel’s back.” Mr. Shields testified that “a sonic boom or an earthquake, even of a minor intensity, could have been enough to create a jar there that might have turned it loose.” They both pointed out, though, that there was nothing in the scientific literature in the field indicating an established causal relationship generally between earthquakes and landslides. Dr. Pipkin, on the other hand, on the basis of his study of the Portuguese Bend landslide and the February 9, 1971, earthquake, opined that there was fairly good correlation between earthquakes and landslides. He further explained that at the time of the June 1967 earthquake the cliff of silty sand and diatomaceous shale on which plaintiffs’ home stood had become so unstable by reason of the subterranean water removing the slight adhesive materials and lubricating the supporting surfaces that a rather slight jar from an earthquake toward the unsupported face of the cliff could cause a landslide.5
Discussion
Defendant’s Appeal
1. The Burden of Proving the Proximate Causation of Plaintiffs’ Loss.
The trial court instructed the jurors in effect that defendant had the burden of proving to them that plaintiffs’ loss was proximately caused by a peril excluded from the coverage of the policy such as earth movement, landslide, etc. In view of this instruction that defendant had to prove its nonliability under the policy, the trial court took the unusual but thoroughly logical step of directing defendant’s counsel to open and close the argument of counsel to the jury on the liability issue. Plaintiffs’ counsel, throughout his argument on this issue to the jury, stated at least five times, in effect, that defendant had the burden of proving to the jurors that the earthquake did *504not cause the landslide. Following about IVi hours of deliberation, the jury requested that the trial court again instruct it on this issue. This time the court told the jury, in effect, that the burden of establishing that plaintiffs’ loss was not proximately caused by the earthquake was on defendant since defendant had to satisfy the jury by a preponderance of the evidence that plaintiffs’ losses were not covered by the policy, and that earthquake was covered but that landslide was not. After about an hour of further deliberation the jury returned the special verdicts underlying the challenged judgment.
The trial court placed this admittedly unusual burden of proof on defendant, as to its liability under the policy to plaintiffs, because of the nature of the policy. The policy is an “all-risks” policy. It is not a specific peril policy, such as a policy of fire and lightning insurance. If it were the latter type of policy and the specific peril covered by the policy had been earthquake, then plaintiffs would have had the burden of proving that the earthquake of June 14, 1967, proximately caused the ensuing landslide. But, here, defendant insured plaintiffs against all risks of physical loss, subject only to certain specific exclusions (including earth movement other than earthquake), and therefore it had the burden of proving its policy’s noncoverage of the loss. This meant that in this case it had to prove that plaintiffs’ loss was not proximately caused by the included peril of earthquake.
Stated otherwise, in an action upon an all-risks policy such as the one before us (unlike a specific peril policy), the insured does not have to prove that the peril proximately causing his loss was covered by the policy. This is because the policy covers all risks save for those risks specifically excluded by the policy. The insurer, though, since it is denying liability upon the policy, must prove the policy’s noncoverage of the insured’s loss—that is, that the insured’s loss was proximately caused by a peril specifically excluded from the coverage of the policy. (See Gaunt v. British & Foreign Marine Ins. Co., Ltd. [1921] 2 A.C. 41, 47, 57-58; Jewelers Mutual Insurance Company v. Balogh (5th Cir. 1959) 272 F.2d 889, 891-892; Annot. 1963, Coverage Under “All Risks” Insurance, 88 A.L.R.2d 1122, 1129-1130; 13 Couch on Insurance (2d ed. 1965) § 48:139, p. 596.)
Applying these principles to the case before us, defendant insurer, in order to establish its defense of noncoverage of its policy, had the burden of proving that plaintiffs’ loss was proximately caused by the excluded peril of earth movement other than earthquake (an included peril). To accomplish this defendant insurer had to prove that the included peril of earthquake did not proximately cause the loss that plaintiffs, the insured, suffered. In other words, it had to negative its exception (earthquake) to its exclusion *505(earth movement) since the burden of proof of its defense of noncoverage of the policy sued on rested on it.6 (Cf. State Farm Mut. Auto. Ins. Co. v. Partridge, supra, 10 Cal.3d 94, 102.) Accordingly we hold that the trial court committed no error in its allocation of the burden of proof on the issue of defendant’s liability and in its instructions to the jury reflecting this allocation.
2. Substantiality of the Evidence.
Defendant contends that the evidence that the earthquake of June 14, 1967, proximately caused the landslide and thereby plaintiffs’ loss, and that the fair market value of plaintiffs’ home immediately prior to this earthquake was $36,410, is insubstantial. Our earlier summary of the relevant evidence regarding the causation of the landslide indicates otherwise. The efficient proximate cause of the landslide that destroyed most of plaintiffs’ home was the included peril of earthquake operating through the excluded peril of earth movement. (See Sabella v. Wisler, 59 Cal.2d 21, 26-27, 31-34 [27 Cal.Rptr. 689, 377 P.2d 889]; Gillis v. Sun Ins. Office, Ltd., 238 Cal.App.2d 408, 415-424 [47 Cal.Rptr. 868, 25 A.L.R.3d 564]; Sauer v. General Ins. Co., 225 Cal.App.2d 275, 277-278 [37 Cal.Rptr. 303].)
We also hold that the jury’s express finding of the fair market value of plaintiffs’ residence just .prior to the earthquake as being $36,410 is supported by substantial evidence. This figure was quite obviously arrived at by taking the square footage of the home from plaintiff Arthur Strubble’s testimony (7,282 sq. ft.), and multiplying that figure by the upper value of $5 per square foot testified to by defendant’s independent real estate appraiser.7
*506 Plaintiffs’ Appeal
The Amount Recovered Is Proper.
Plaintiffs contend that their $36,410 home was a total loss, rather than the partial loss found by the jury of $18,610—the difference between the fair market value of the home before the earthquake and its corresponding value of $17,800 thereafter. In support of this contention plaintiffs argue that their home after the earthquake actually had a negative value instead of $17,800 because it would have cost more to move what was left of the home to safe ground than the remainder was worth on the market.
In finding the value of $17,800 for the remainder of the building, the jury, again obviously, took the figure of defendant’s independent appraiser for the square footage remaining (3,560) and multiplied it again by the upper value per square foot he had ascribed to it..
Plaintiffs attack this remainder value of $17,800 on the ground that the defendant’s appraiser testified that his corresponding figure of $17,000 could only be realized by removing the structure back to stable ground. He did so testify, but his assumption that the ground beneath the remaining structure was unstable was belied by Dr. Pipkin’s testimony and report to the contrary.
The judgment is affirmed. The parties shall bear their respective costs on their appeals.
Ford, P. J., concurred.