Update:
I think wget would be better, i just don't use it very often:
wget -r -A.zip http://dds.cr.usgs.gov/srtm/version2_1/SRTM3/
Also see for some retrieval wait options (if you want to be courteous):
https://superuser.com/questions/178297/wget-terminates-unexpectedly/178326#178326
You could curl the files with a little bit of text editor work.

For curl, the man page is here:
http://curl.haxx.se/docs/manpage.html
Download here: http://curl.haxx.se/download.html
I think there are throttling and retry options (check the above-mentioned man page).
curl -O http://dds.cr.usgs.gov/srtm/version2_1/SRTM3/Australia/S11E119.hgt.zip
curl -O http://dds.cr.usgs.gov/srtm/version2_1/SRTM3/Australia/S11E120.hgt.zip
curl -O http://dds.cr.usgs.gov/srtm/version2_1/SRTM3/Australia/S11E121.hgt.zip
curl -O http://dds.cr.usgs.gov/srtm/version2_1/SRTM3/Australia/S11E122.hgt.zip
curl -O http://dds.cr.usgs.gov/srtm/version2_1/SRTM3/Australia/S11E123.hgt.zip
In addition to Jay's excellent pointers, I would recommend using the --timestamping flag when downloading so you can only get files that are newer than on your file system (really useful when your downloading gets interrupted).
Additionally, using the --no-parent flag means that only the SRTM3 directory is downloaded rather than it continuing up the tree to the other directories.
The full wget commandline I'm using:
wget -r --no-parent --random-wait --timestamping -A.zip http://dds.cr.usgs.gov/srtm/version2_1/SRTM3/