Crap - I refreshed instead of hitting post, all my typing got dumped.
Anyway,
Here's the odd part - I had this working, so I went thru and commented out the various Serial.print's that had been added for debugging, as they impact how big the array can be. And it stopped working! So I put them back in thinking maybe a timing thing - no help.
The processor is just a '328, actually using a Duemilanove card.
I have
const unsigned int arraySize = 1200;
byte displayArray[(arraySize-1)];
defined in the pre-setup code area so I can change the size in one place easily.
I've had the array at 1500 before I started the EEPROM storge, had knocked it down to 1200. Can try a little smaller.
Is odd to me that the two addresses are apparently being read back properly before addresses 2-1199 are written & read, and then 0-1 cannot be read.
x=x+1. I just like being able to quickly see what I have going on. You might say I don't completely buy in to the C++ abstractions & overloading & structures vs functions & whatever - I just treat it like fancy BASIC and x=x+1 suits my needs. If I could, I would still use "next x;' 8)