Plaintiff is a common carrier transporting highway freight in trucks and trailers. A considerable portion of plaintiff’s trailer equipment was purchased from defendant, a corporation engaged in the manufacture and retailing of such equipment. In August of 1944, plaintiff purchased from defendant two semitrailers. Such trailers are equipped with rear wheel's only, the forward end of the trailer resting upon the rear end of the tractor that furnishes the motive power to pull such equipment. The evidence shows that underneath the forward part of the trailer, and on the rear of the tractor, there are metal bearing plates where the two come into contact. These bearing plates are called “fifth wheels. ’ ’ In order to hook the trailer to the tractor a so-called adaptor pin or kingpin is used. This pin is a substantial piece of equipment, and fits into slots cut into the respective fifth wheels. Various kinds of pins are used in different equipment. If both pieces of equipment are full automatic, the kingpin is a fixed part of the tractor’s construction, and when both are semiautomatic the kingpin is a fixed part of the trailer’s equipment. But when a full automatic is connected with a semiautomatic they are joined with a detachable kingpin which is then called an adaptor pin. The evidence shows that when plaintiff purchased the two semitrailers defendant was informed that the hookup was to be automatic to semiautomatic, that the parties discussed the type of adaptor pin to be used, and that defendant advised plaintiff as to how this pin should be used. Such a pin was purchased from defendant by plaintiff at the time of the purchase of the two semitrailers, at which time one semitrailer was delivered to plaintiff.

*622 The two fifth wheels have slots cut into them to fit the kingpin. The pin itself is fitted into a piece of boiler plate about 10" long and 1" thick, and of irregular shape that fits into the irregularly cut hole of the fifth wheel. Through one end of the boiler plate a hole is bored and into this hole the pin is pressed. The pin is of very hard steel. On one end it is 2yz" in diameter with a flange left on as an integral part of the pin. That is the upper side of the pin which projects upward from the boiler plate about 2y%". The lower end of the pin projects below the boiler plate about 2" and is machined to 2" in diameter. At the outer end of the lower side of the pin is a flange of soft metal, being a ring %" thick and %" wide. This soft metal flange is welded to the hard metal pin. The upper flange keeps the adaptor pin from falling through the upper fifth wheel, and the lower flange keeps it from bouncing out of the lower fifth wheel. The causes of action here involved are based upon the alleged failure of this adaptor pin and flange to perform the services required of them.

Shortly after purchasing the pin and accepting delivery of the first trailer from defendant, one of the drivers of plaintiff had an accident in which the new semitrailer tipped over. After the accident it was discovered that the soft metal flange on the new adaptor pin had broken off. Plaintiff’s president immediately discussed the matter with the branch manager of defendant. Plaintiff’s president then expressed fear that the adaptor pin was not safe to use. The manager, after conferring with other officers of defendant, assured plaintiff’s president that the pin was safe and stated that he would guarantee them. Under these circumstances plaintiff purchased from defendant another adaptor pin and accepted delivery of the second semitrailer. It was on the first round trip that this equipment was used that the accident occurred that forms the basis of this action.

The accident occurred at about 1:30 a. m. on October 1, 1944, in Oregon. At that time plaintiff’s tractor, weighing 6 tons, was drawing the new semitrailer connected with the new adaptor pin. The trailer when empty weighs 7 tons, and at the time of the accident it was loaded with 15 tons of cargo. The truck driver, Olin, was thoroughly familiar with the route, having driven it many times. On the night in question he had been driving about seven and one-half hours before the accident, and, up to the time of the accident, had *623 had an entirely uneventful run. The accident happened just before the equipment entered upon a bridge crossing the South TJmqua Creek in Oregon. For some distance before the bridge is reached the highway is relatively level, is smooth, is two-laned, is paved with asphalt, and is 26 feet wide. There was no.other traffic on the road at the time. A short distance before the bridge is reached there is an upgrade, and just before the bridge is reached there is a sharp left turn still on the upgrade. The bridge, also on the upgrade, is but 19.3' wide, this width being narrower than the highway.

Olin testified that the maximum speed of his tractor on the level was 43 to 45 miles per hour; that prior to the accident, and while on the level, he had been in the highest speed; that when he hit the upgrade he was going about 35 miles per hour and that he then shifted gears to a lower driving speed; that he approached the left-hand curve at about 25 miles per hour; that he was on his side of the road and took the curve at the speed he normally drove around that curve; that it was a normal operation “the same as every night”; that just before he entered the bridge the tractor spurted or lurched ahead with accelerated speed; that he looked through the mirror and saw that the trailer had disappeared. He stopped and discovered that the detached trailer had rolled down the embankment. It was still on its wheels but the sides and top had broken open. The trailer and its load were practically a total loss. Olin examined the fifth wheel of the trailer and discovered that the soft metal flange had torn off the adaptor pin. He later found the flange.

Based upon these facts the plaintiff brought the present action. The complaint is in two counts, one on the theory of an implied warranty and the other on negligence. Both causes of action are based on the theory that the trailer became detached because of the failure of the adaptor pin. The cause was tried before the court without a jury. The trial court found in favor of plaintiff on both causes of action and judgment was entered in favor of plaintiff for $14,665.54.

On this appeal defendant’s basic contention is that the evidence is insufficient, on either the warranty or negligence theory, to sustain the finding that the failure of the adaptor pin was the proximate cause of the accident. At the trial it was defendant’s theory that the flange was torn off because the driver took the turn too fast and tipped over the trailer, *624 and that in normal operations no strain at all is placed on the lower flange. Defendant now urges that plaintiff failed to prove that the failure of the flange was the proximate cause of the accident.

It is elementary law that on such an appeal this court cannot disturb a finding of fact unless there is no evidence or no reasonable inference from the evidence to sustain that finding. (Crawford v. Southern Pacific Co., 3 Cal.2d 427 [ 45 P.2d 183 ]; Dawson v. Boyd, 61 Cal.App.2d 471 [ 143 P.2d 373 ].) The above rule applies to a finding of proximate causation. (Matsuda v. Luond, 52 Cal.App.2d 453 [ 126 P.2d 359 ].) Of course, the inference or inferences indulged in must be reasonable, must be based on the evidence, and cannot be the result of mere guess, surmise or conjecture. (Reese v. Smith, 9 Cal.2d 324 [ 70 P.2d 933 ]; Puckhaber v. Southern Pacific Co., 132 Cal. 363 [ 64 P. 480 ] ; McKellar v. Pendergast, 68 Cal.App.2d 485 [ 156 P.2d 950 ].)

Tested by these standards we are of the opinion that the reasonable inferences from the evidence amply support the finding that the failure of the soft collar flange on the adaptor pin to sustain the load placed upon it in a normal road operation was the proximate cause of the accident. Most of the evidence on this issue was that of experts. In addition, the adaptor pin and the torn off flange were introduced into evidence, and the trial judge examined not only the exhibits but the fifth wheels into which the adaptor pin fits.

The plaintiff’s experts had examined not only the broken flange and pin, but had examined other pins manufactured and sold by defendant. These were also introduced into evidence. All the exhibits have been transmitted to this court pursuant to rule 10(b) of the Rules on Appeal. Plaintiff’s experts testified that the purpose of the soft metal flange on the lower end of the pin was to prevent the pin from bouncing out of the fifth wheel and from slipping out of the connection. All of the experts agreed that on a level smooth road no force would be exerted vertically on the pin, all the force then being a horizontal pull. It is quite clear to anyone, however, that with various road conditions, upgrades, downgrades, curvature of the road, road irregularities, turns, etc., pressure is exerted on the pin in a vertical or partially vertical direction, the amount and extent of such pressure varying with the conditions. Plaintiff’s expert Dresser stated that from his examination of the pin, the scratches thereon, *625 and of the distortion of the broken flange, he would say that approximately a vertical force was exerted against the pin and flange and that this caused the flange to be forced off. The broken flange was so distorted that it indicated to him that a sort of prying action had forced off the pin. The flange, according to this witness, was forced off because of the failure of the weld by which it was attached to the pin. The weld was on the lower or outside portion of the flange. Dresser opined that the weld would have been much stronger to withstand a vertical or a prying force had it been on the inside or upper side of the flange, and would, of course, have been very much stronger had it been welded on both sides. He gave it as his opinion that “the way it was done is the weakest way that it could be done.” Plaintiff’s engineer King stated that the weld was “a very light weld” and that the method employed by defendant to weld the pin was “very very poor. ’ ’ To him the distortion of the collar demonstrated that there was a “rocking” force exerted upon the pin and flange from both sides. The scratches on the pin and flange demonstrated to him that there had been a progressive fracturing of the weld and that some final strain forced it off. He opined that with a sufficient weld the welded portion would have been as strong as the metal itself. He also gave it as his opinion that the marks on the pin and collar could not have been made by the trailer capsizing first, because, if that had happened, only one side of the pin would have been marked, while here both sides were marked, indicating a rocking force exerted on both sides; that by the laws of physics the centrifugal force exerted in rounding a curve is always exerted towards the outside and not the inside of the curve, and here the pin was scratched on both sides.

Defendant’s testimony conflicted with some of the above. Its manufacturing manager testified that the company had manufactured and sold some 2,000 such pins, and that to his knowledge the only pins that had ever failed in operation were the two sold to plaintiff. The purpose of the collar, according to him, was to keep the pin from bouncing out of the fifth wheel, and, unless the trailer capsized, no appreciable force would be exerted against the flange. The weld, to him was “a good average weld.” When the trailer was empty there would be a bouncing force on the pin, but with a loaded trailer on a normal road operation there would not be enough bouncing force to disengage the pin. Expert Wood- *626 ruff, called by defendant, thought' the weld was “of better than average quality,” and he was of the opinion that the flange was twisted off by such a force as would have been exerted if the trailer had capsized.

This is a fair summary of the evidence. Interpreted most favorably in favor of plaintiff, it shows that the trailer became disengaged from the tractor; that the adaptor pin which holds the two pieces of equipment together became disengaged; that the flange which keeps the adaptor pin from becoming disengaged was broken off the pin; that the flange was attached by a very weak weld, and that several practical means had not been adopted to make that weld stronger; that, had the weld held, the two pieces of equipment could not have become disengaged; that, had the weld held, the trailer could not have capsized without tipping over the tractor or tearing out the fifth wheel assembly; that the tractor did not tip over, and that the fifth wheel assembly did not rip out; that the two vehicles became disengaged during a normal road operation; that when they became disengaged the tractor spurted forward—not to the side as it obviously would have done if the trailer started to tip over; that the flange was forced off by a rocking motion exerted on the pin; that a rocking or prying force would have been exerted against the pin and flange by a heavily laden truck going around a curve. From that evidence it is certainly at least a reasonable inference (if not an inevitable one) that the two pieces of equipment became disengaged as a proximate result of the pin becoming disengaged during a normal road operation, and that such pin became disengaged because of the failure of the welded flange to sustain the normal load intended to be exerted upon it. This being so, the finding of proximate cause is amply supported. This sustains the judgment on either theory here pleaded and found.

The judgment appealed from is affirmed.

Ward, J., and Bray, J., concurred.