The tail end of the story is not as funny as the rest.  The impact gun tests settled that the problem was not related to impact damage.  I let the chemists worry about obtaining a selection of long chain fatty acids to play with and test against the chemistry of their coatings.  I figured that people who knew about chemistry would be better suited to testing them out.

Our windows were flying again and the coatings seemed to be doing better.  But in a meeting a couple months later I realized that no one was taking up the gauntlet of fatty acid testing.  I twisted the arm of one of the chemists a little bit and he showed up in my office with a chemical reagent catalog.  He flipped through and identified the seven long chain fatty acids in the twelve to eighteen carbon range referred to by the entomologist I had talked to so long before.  I wrote out a purchase requisition for them and sent it off.

They arrived in the form of bottles of dry powder.  I turned them over to my chemist friend hoping he would take it from there.  He gave me back bottles of liquid acid prepared from each powder.  So I pulled some old samples from the cardboard boxes that cluttered 1/3 of my office space to test them out.  (I learned never to discard samples until their projects were well cold in their graves.  It made for a cluttered office, but a good reputation.)  I exposed several coatings with known bug reactions to the chemicals.  Some chemicals left marks, other did not.  I selected three that were promising.  They tended to mark coatings that I knew were vulnerable but did not damage ones I knew were okay.

The acids I had did not truly duplicate the effect of the real thing, but at least they seemed to screen for the
right properties.  To try and exactly duplicate damage done by unknown chemistry in unknown conditions is really to much to hope for.  Sometimes having a quick and simple test that is close to real is much better than having a perfect test that is very tricky and takes a long time.  I gave the acids back to the chemists.  The bug juice simulation seemed to work well for the chemists and they used it to screen new coatings with good success.

So the story ends on the day when my boss asked me to write up a detailed procedure for "Bug Juice Resistance Testing" which he could include it in a new test program for an advanced bomber windshield.  Somewhere out there in the industro-military complex are technicians working to my instructions. They have no idea of my cricket gun or what uses I put to that box of Raisin Bran that I carried into work one day.

I suppose the moral of the story is that expertise in the area of a problem is not the prime requisite for problem solving.  Problem solving ability is the prime requisite, and expertise is something you develop or seek out as part of the process.  I can only hope that all the other dull test procedures I had to learn while in the test lab have such colorful backgrounds as my own contribution.