The City of Ballwin charged defendant with speeding, driving 50 miles per hour in a 40 miles per hour zone. On defendant’s appeal to the Circuit Court, he was found guilty and fined $10 and costs. On this appeal, defendant challenges the evidence concerning a radar gun used to determine his speed, contending it was inadmissible since the radar gun’s accuracy was not shown by first establishing the accuracy of any of the devices used to test the radar gun.
Officer Townsend testified he used the radar gun after letting it warm up for three or four seconds before testing it. He stated he tested the machine by means of its “internal calibration equipment,” not described in the record. The officer then tested the radar with a tuning fork. Following that he had Officer Kirk drive a car through the radar gun’s range. In the officers’ opinion, all three tests indicated the radar gun was operating accurately, but there was no showing of accuracy of any of the three “measuring devices” used to check the radar gun.
We endorse the statement in State v. Graham, 322 S.W.2d 188[18] (Mo.App.1959), that the value of any test of radar speedmeter depends upon the accuracy of the measuring device against which it is cheeked. We also share the opinion in City of St. Louis v. Boecker, 370 S.W.2d 731[3] (Mo.App.1963) that the particular measuring device must itself be shown accurate before the radar evidence is admissible. Here, Officer Townsend made what plaintiff contends were three different tests of the radar’s accuracy. However, not one of the three tests — the internal calibration equipment, the tuning fork or Officer Kirk’s patrol car speedometer — was itself tested for accuracy.
The conclusions to be drawn from the three radar speeding cases in Missouri, State v. Graham, supra, City of St. Louis v. Boecker, supra, and Kansas City v. Hill, 442 S.W.2d 89 (Mo.App.1969), are that the accuracy of a measuring device must be suffi*281ciently shown before the radar reading is considered accurate. Kansas City v. Hill, supra, requires the existence of a dual test before a court can accept radar speedmeter evidence.
Here there is only one test of the radar device’s accuracy in evidence. The tuning fork, when struck in front of the radar gun, caused the device to register 65 miles per hour but the accuracy of the tuning fork was not shown. There would have to be a second test before the trial court could accept the accuracy of the radar gun. We do not find a second test in the record.
Although it is stipulated the internal calibration equipment was activated, registering 40 miles per hour on the radar gun, we are not told whether this reading was accurate. We have no foundation on which to base an assumption that the internal calibration equipment demonstrated the radar was working properly.
Officer Townsend, operating the radar gun, stated the gun registered 40 miles per hour on a drive-through test by Officer Kirk in his patrol car. Officer Townsend testified Officer Kirk radioed he was traveling at 40 miles per hour as he passed the gun. Officer Kirk recalls driving past the radar gun at 35 or 40 miles per hour, but the car’s speedometer was not shown to be accurate, nor was it compared with either of the other testing devices. The proponent of the radar evidence has the burden of proof as to its accuracy. City of St. Louis v. Boecker, supra [2]. We think the evidence on these two tests equivocal at best. Plaintiff has failed in its burden. The present case contains no other evidence as to the speed defendant was traveling besides the inadequately tested radar reading of 50 miles per hour. Here, unlike State v. Graham, supra, there was no observation by an experienced traffic police officer that defendant was exceeding the speed limit, nor admission by defendant that he had been speeding. Officer Townsend could not testify as to any personal knowledge of the accuracy of any one of the three measuring devices, as did the arresting officer in Kansas City v. Hill, supra.
Without evidence of the accuracy of any one of the measuring devices or any independent observation of the speed of defendant’s car, the plaintiff has failed to meet its burden of proof.
Judgment reversed.
KELLY and STEWART, JJ., concur.