This sounds a lot like what happened to me, when doing a Win2K + Win2K
dual boot on the same drive. Second OS relied on the first, for the
files used to boot. And the system drive letter ends up being "E:".
Note that the Microsoft terminology is reversed. "(System)" holds
the boot files. While "(Boot)" holds the vast majority of the C: files.
A single partition can perform both roles at the same time. But
if circumstances permit, the installer is allowed to split them.
Looking at the picture, it may have spotted
1) Boot flag on the 426.70GB partition.
2) Files like boot.ini on the 426.70GB partition, implying
support for booting. The installer, can simply add a line
to the boot.ini, to add the "new OS".
Also, note how E: is a logical. Which is a way of saying
loading process, expects the boot flag to be on the partition with
the bootable files. I do not think it makes sense, to set the
boot flag on an Extended with Logicals. (Note that Linux has no
restrictions on installing the OS on logicals. It does not even need
the boot flag to be set. it is the MBR code, that decides these things.)
If I was attempting to fix it, I would.
1) Boot a maintenance OS (or connect the disk needing maintenance, to
another computer temporarily).
2) Delete E;, the failed install. (You cannot really do that, unless
you are using some other OS to do it, or connect the drive to another
computer to clean it up.) Remove Extended partition and logical partition.
3) Remove the boot flag from the 426.70 GB partition. You can do
that with PTEDIT32.exe . Or via Linux LiveCD and "sudo fdisk" type command.
4) If the "BBC_docs" implies the 426.70GB is not a system partition,
remove the system-like files (old copy of pagefile, hiberfil, boot.ini
and so on, all the cruft put at the root level for an OS install). That's
to reduce the odds the Windows installer will misinterpret what it is seeing.
5) Using PTEDIT32, move the 426.70 GB partition table entry, to the
second line in the partition table. This leaves room in the first line,
to define the "fresh install" partition.
6) Add a partition, FAT32 or NTFS, for a new "left-most" partition for
Disk 0. Verify, with PTEDIT, it is the first entry in the table.
Preferably, the entries in the partition table, should be in spatial order.
It makes it easier for humans to understand the set up later. (I had a
machine with the entries in the wrong order, and it used to drive me nuts,
when I had to fix something, and the entries did not make sense. And I had
to keep telling myself, which was which. Making them in spatial order
now, pays dividends later when you are in a panic.)
7) When the install starts, tell it to use that new "left-most" partition
for the space it needs. The purpose of making that partition first, is
to raise the odds it becomes "C:".
The purpose of that, is so the next install attempt, does not do the same thing :-)
Just some ideas,