Jim Smith wrote:

Isn't this in the wrong NG??

No. I admit to being way ahead of my time on most things but this case
is about the pipe that must have existed that conducted the sound
through the atmosphere. (Assuming it came via the atmosphere.)

There would have been massive cyclonic activity concurrent with the
series of events -if that helps. They would have been in the shadow
zone, a band roughly 15 degrees either side of an arc at 120 degrees
around the epicentre.

Dates to check:

Beginning 20 May 1883, three months before the final explosion, steam
venting began to occur on a regular basis.

By early August, three vents were regularly erupting on Krakatoa; tides
in the vicinity were unusually high,

11 August saw the onset of larger eruptions, with ashy plumes being
emitted from as many as eleven vents.

On 24 August, eruptions further intensified, and the cataclysmic phase
began on Sunday 26 August, near midday.

The 27 August eruptions occurred at 5:30 a.m., 6:42 a.m., 8:20 a.m.,
and 10:02 a.m. local time.

By 28 August, Krakatoa was quiet again. Ash clouds caused by the
eruption blocked sunlight for a couple of days within the area.

http://en.wikipedia.org/wiki/Krakatoa#The_1883_eruption

Nice of it to stay in the season of North Atlantic hurricanes. That
should make tracing records easier than they might have been.

Well, that's all from the genius; now onto the clerks and number
crunchers. And/or sundry midgets. No offence. I couldn't do your job. I
I respect your abilitiies but well, no...

I keep forgetting to be humble.. sorry.