what is res.on() in http request?

How does the callback in an http request work? Why is there a second callback? I thought the callback supplied with the request would be called when the response was returned, yet this seems to be placing another callback waiting on... something...?

var http = require("http");
http.get("http://www.espruino.com", function(res) {
 res.on('data', function(data) {
  console.log(data);
 });
});

When does res.on() get called?
Does res.on('data') get called only once per request, or can it be called multiple times, each time with a piece of the data? If the latter is true, how do I know that i've got all the pieces?

The link in the http docs for more info on the node.js http api is dead - possibly when working this would explain it - but nodemanual.org seems to be down. http://nodemanual.org/latest/nodejs_ref_­guide/http.html

The use case I've got in mind is - I want to fetch a file and evaluate it, and am wondering if I can safely do this, or whether sometimes the data won't be complete, hence the eval will fail.

var http = require("http");
http.get("http://master/program.js", function(res) {
 res.on('data', function(data) {
  eval(data);
 });
});