Geminid passing through Orion's shield.
Exposure: 800 ISO, 26 seconds, F2, 28mm lens
Most of my 35 photos between 4 and 5 am were 1 minute long, the exception being that when I saw a meteor in or near my field, I'd stop my exposure so the background doesn't get any brighter. My photos represent about a 1/2 hour of cumulative exposure. Catching six or seven out of 35 minutes suggests a rate near 10 per hour. If my camera's field of view were 1/10th of the sky -- I'm confident its field is smaller -- that would suggest that the predicted rate of 100 per hour under a moonless sky was accurate. Caution should be taken with this estimate as it's based on my small sample group.
Why do I keep saying "six or seven"? Because one photo has three streaks but two are so closely in line that they may be one meteor:
Three faint streaks can be seen in this photo of Coma Berenices.
Tracing the streaks reveals that two are on the same radiant.
I've photographed two meteors apparently on one radiant during the Leonids in 2001, but this was on slide film and the two meteors were far apart on the film. So, I expected some distortion: that is, the lens would introduce some distortion and the film's slide mounting (lack of flatness when scanned) would introduce more.
Two or one meteor passing near the Pleiades in Taurus.
Tracing reveals parallel paths that could be on the same radiant.
Curvature of the slide film can be seen at the bottom center of the frame.
I never believed that I had captured a meteor appearing, then dimming, then reappearing, but I think I have now.