While investigating #1060 I found that the aerogel material was constructed with uninitialised 0th element in the arrays RayleighPhotE[] and RayleighSL[]. I don't know if this can affect physics, but I set the 0th element to 0. so that at least results would be consistent (previously took random set of 1s and 0s into this value).

  a1_mt->AddProperty("RAYLEIGH",     RayleighPhotE, RayleighSL,63);

Will be in 0.3.2