Chris and Roger,

I took a look at the numbers, and plotted them myself a couple of ways.

First, the general character of the aftershocks versus distance averaged to 5-10 degree scales -

You can probable guess how the distribution should look:
if aftershock occurrence were not affected by prior earthquakes, the probability would be proportional to the area of the Earth's surface at that distance, or the sine of the distance. So growing linearly for a while, then leveling off at 90 degrees, and falling back towards 0 at 180 degrees. To first order, that is what the curve does.

The exception is close in, within 20 or 30 degrees of the earthquake there is more activity than would be present at random. According to my prejudices, only 0-5 degrees or so should show such a bump, so there is a suggestion of triggered activity out to surprisingly large distances. It may be, however, that earthquakes tend to be in seismically active zones, which might produce a similar curve, so a better test would be to compare this distribution with the same set of earthquakes at random times, to separate the spatial pattern of active areas from the triggered aftershocks.

Then the fine structure -

The only anomalous feature is the peak at 101 degrees, with 9 events, and 102 with a moderately high number. However, 103, 104, and 105 degrees have very FEW events.
The rest of the peaks fall with a reasonable distribution, although there's a bit of a shortage of bins with 3.

So there is some suggestion of something at 101, although it is not compelling, and the 0-20 excess could be investigated.

John