NScan.org: the legend is still here ;)
Traceroute

Traceroute is designed for quick network route discovery. Unlike normal traceroute, it traces all the way to host at once and saves tenth to hundreds times. It also measures the time necessary for packet to return and looks up all intermediate routers. Unlike normal tracert it's based on UDP protocol, and allows to trace networks where incoming ICMP messages are filtered.

The way it traces the route makes time measurement to be inadequate, but it can be used as relative values. Assume first hop to be <10 and subtract the difference from other hops. Please note that far hop may take less time than near - it can be also seen in ICMP traceroute, used by Windows. It's really so - it's a time, lost for internal router processes before it answers.

Interface details:
Tries: maximum retry number if router was not discovered. Traceroute will trace the route again until exceeds this value.
Hops: maximum number of hops to trace. This is assumed route length. Usually do not exceed 30 hops and can't be more than 255. Be careful with the hop number - too large number will flood the target host.
Time (milliseconds): maximum time to wait for router's answer. Timeout assumes that the packet was lost.
Button "Host >" copies the clipboard contents to the question line.

Operation description:
When started, traceroute attempts to trace all the way at once. Usually it takes about 2 seconds to do it. If some of the routers were not discovered it waits until timeout and re-asks the route. Upon completion it narrows the results to the route length and looks up all the intermediate routers.