an example: to use a cdrom in windows, the drive is always there, with a letter. when you put a disc in, it makes the data available through that letter. In linux, instead of D:, you would usually have a DIRECTORY called /mnt/cdrom, and when you put a CD in, if it isn't set to mount automatically, you would type "mount /mnt/cdrom".
Take a look at /etc/fstab. note the filesystems listed there. if you want to mount something not present there (say you create a file that has its own filesystem, or want to use an ISO file without burning it to CD - this is called "loopback") you would use the following syntax:
mount -t <filesystem_type> <path to file> <mount point>
the filesystem_type could be ext2 or ext3 (linux), ntfs (duh), vfat (fat32), msdos (fat16) and there are a few others you wouldn't normally use. the path to the file is just the location of the file. the mount point is where the new filesystem is going to appear. You can mount it to ANY existing directory, such as /tmp, /bin, /usr, /mnt, etc. it is just a very bad idea to mount something over a directory that has data of its own, since you will have to unmount to access the existing data. if you mount your mp3/pr0n/movie hard drive to /bin, you won't be able to access anything that was in /bin before. I tihnk that means you wouldn't be able to unmount the partition. If data there is in use currently, it may give you an error and not mount.
I can't try this now since my linux boxes are off for the summer.