well first off, unix has the design concept fo a "Unified Filesystem", which means that "everything is a file(tm)" this means the /dev/hda1 etc are just points of access to the partitions and while they're supposed to represent the filesystems, they're just files inside the larger system.
what this means is the everything gets put into one filesystem for simplifying the whole deal (rather than the VMS-style A: B: C: D: representation windows uses) this is so you won't have to know what partition a file is on, just where it is in the filesystem, which gives you a lot of freedom on partitioning if you know it won't break anything.
the /dev/hda hdb, etc are access points which means if you try to read them, you get the raw data on that partition, which is usually only useful for either mounting a partition or formatting one, regular users/admins hardly have use for it.
as an example: say hda1 was your root partition, but you wanted the hdb2 partition to be used for your users' home directories. Since all home directories are located under /home, you just mount hdb2 to /home and anything written to a file underneath /home will actually be written to the second partition on the second IDE drive(hdb2), instead of somewhere on first partition on the first drive (which is called hda1)
for more info:
http://www.tldp.org/HOWTO/Partition/
http://www.pathname.com/fhs
I hope this was helpful
