SCSI sick on 7300

[Bruce McIntyre <bamc2@student.monash.edu.au>]

Hello,
When attempting to install Debian (2.2.17 and .19 so far) at various 
stages in the install process as well as on restart ( if I was lucky 
enough to get that far) I have been getting many SCSI errors.
I have a Quantum Fireball TM2100s Rev 300Z which i am attempting to 
dual boot with BootX ( later I want to ditch MacOS  )
I have tried the drive on both Id 0 and 1, as well as on the Mesh and 
the 53C94 controllers.
When attempting to install kernel, or base, (while in the rescue 
floppy or booting from the BootX image) I recieve nasty stuff like:

SCSI disk error: host 0 channel 0 id 0 lun 0 return code = 8000000
scsidisk I/O error; dev 08, 07 lun sector 262168
    read_inode_bitmap : Cannot read inode bitmap
    ext2_check_inodes_bitmap: Wrong free inodes count in super block, 
stored = 217717, counted = 202165

I  have tried many different partition schemes; some even got me as 
far as starting emacs, but at every startup fsck would find more 
problems and files would just dissapear to lost_and_found....

thanks for any ideas about this...