Affirmed.

ON PETITION FOR REHEARING

Appellant asks us to reconsider our affirmance without opinion, suggesting that we have applied here a different standard from that applied in Barry v. Greyhound Lines, Fla.App.1968, 216 So.2d 775. There we upheld the sound judgment of a trial judge who admitted testimony of measure*228ments of speed of vehicles passing through an intersection marked by a caution signal and to which no specific statutory standard applied. That the trial judge best knows whether proffered testimony of this sort tends to clarify or obscure the issues is a factor too little regarded by appellate counsel, and often lost when a court’s language is excerpted into headnotes. See Continental Casualty Co. v. McClure, Fla.App. 1969, 225 So.2d 590. Here the trial judge excluded tabulations of speeds at which motorists crossed the railroad tracks at which plaintiff’s husband was fatally injured when defendant’s train struck his truck.

There is one major difference. In Barry, the bus had the right of way and the average speed was suggested as a standard which the jury might attribute to a reasonable man. In this case, White was under a duty to yield to the train. Plaintiff’s able counsel, perceiving this difference, sought to introduce the testimony as bearing on the railroad’s knowledge of the speeds at which motorists cross the tracks. We think that exclusion for this purpose was sound. Certainly there is no continuing legal duty on the railroad to make such measurements as the plaintiff did in order to determine whether the crossing is sufficiently dangerous as to require an automatic signal. This argument was otherwise raised and determined against the plaintiff.

Furthermore, the statistical method employed in this case was erroneous. The data were tabulated within ranges of speed, including a range from 30 to 50 miles per hour which embraces the 42 miles per hour plaintiff contended White was traveling, but included an indeterminate number of considerably more prudent drivers.

While in Barry the incidence of specific speeds clocked across the intersection was not in evidence, there was no apparent statistical fallacy. We repeat that the sound judgment of the trial judge is the primary factor to which appellate courts must look in evaluating testimony peripheral to the main issues and tending to prove them indirectly. In the absence of a showing that the trial judge’s judgment is demonstrably wrong, we are obligated to affirm.

The petition for rehearing is denied.