Maxwell 37667 dies interestingly...thoughts?

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: "J. Aaron Holmes" <jaholmes@xxxxxxxxxxxxxxxxxx>

I took my Maxwell 37667 to a friend's place over the
weekend to see if it would help get his new 24" coil
to put out some sparks.  Despite early indications
that things were too far from complete to run the
thing, we got enough wire stripped quickly enough to
do a temporary lash-up before daylight failed.

Things kicked off with the destruction of two homemade
rolled caps that he'd spent a bunch of time on.  What
a bummer (and boy did they look cool, too!).  I took a
deep breath and offered to put the Maxwell on the
cooker to see what would happen.  We were immediately
gratified with 8- and 10-foot streamers as the break
rate topped 400 bps.  Yikes!  We dropped a few
different toploads on the thing and played around for
a while.  In the end, the Maxwell probably had only
two minutes of runtime spread over a ten-minute
period.

Then, Bad Thing #1 happened:  A streamer hit the wire
between the Maxwell and the primary.  Due to the
temporary nature of the wiring, the cap had been
sitting relatively exposed a few feet from the base of
the coil.  Doh!  A metal sawhorse was quickly placed
over it.

And finally, Bad Thing #2:  Shortly into the run
immediately following placement of the sawhorse, a
crack opened in the Maxwell, and the oil drained out
rather quickly, leaving a nice puddle of Maxwell juice
on the floor (VEERRRYY slippery stuff, I can now say!)

What is quite interesting about all of this is that,
due to the surpise of seeing the Maxwell croak, it
took us a few seconds to kill the power.  In those few
seconds, the coil continued to operate normally; there
was no obvious dip in performance.  The contents of
the Maxwell were quite warm, though not what I'd call
"hot", and I neither saw nor smelled any sign of
burning.  On top of that, measurement revealed that
the capacitance was still 0.03uF, and no DC resistance
would register on the DMM.

So the mode of failure here is a little unclear to me.
 Did thermal expansion of the contents simply pop the
case, or was there most likely some kind of arcing
fault, perhaps initiated or encouraged by the strike
in the prior run?

As the Maxwell's guts seemed in remarkable shape, my
friend is going to try running them in a bucket of
mineral oil and see what happens.  I'll report back
with the results.  If they aren't serviceable, then
perhaps we'll pry them apart even more and see what we
find.  They were just too pretty not to get a second
chance :-)

Any other thoughts on "Why Maxwells Fail"?  I never
even felt the thing get warm in my little coil running
at 120 bps.

Regards,
Aaron, N7OE