Historical hurricane tracking

I was just looking through the Unisys Atlantic Hurricane archives, when I began wondering-- how in the world were hurricanes in the mid-1800s tracked? For example, look at "Hurricane 3" in 1853 at http://weather.unisys.com/hurricane/atlant...853/3/track.dat ... How do they have 6-hrly lat/lon, max winds, and central pressure data for this hurricane (from 1853!)? I assume nearly all of it may be based on local documentation -- a huge storms comes ashore, and the citizens write about it. But how does that work with storms that don't ever come ashore? I'm sure there have been ship reports, but I still find it odd that they can track these hurricanes that well... News writings + ship reports + climo ? Just curious...
When I was at Creighton there was a PhD candidate focusing his research on piecing together archived shipping logs to reconstruct the paths of hurricanes in the late 1800's / early 1900's. I wouldn't be surprised if his work has been used to further substantiate the accuracy of these hurricane climatologies.