question on NFA and DFA generated by flex

Consider the following simple flex input:

++++++++++++++++

var v

%%

{var} {}
.. {}

%%

int main()
{
  yylex();
  return 0;
}

++++++++++++++++

Running flex in trace mode (flex -T scanner.l) it tells me
about the generated NFA and DFA:

++++++++++++++++

%%
1       (v)
2       .
3       End Marker


********** beginning dump of nfa with start state 11
state #    1    257:     0,    0
state #    2    257:     0,    0
state #    3    118:     4,    0
state #    4    257:     0,    0  [1]
state #    5    257:     1,    3
state #    6     -1:     7,    0
state #    7    257:     0,    0  [2]
state #    8    257:     5,    6
state #    9     -2:    10,    0
state #   10    257:     0,    0  [3]
state #   11    257:     8,    9
********** end of dump


DFA Dump:

state # 1:
        1       4
        2       5
        3       6
state # 2:
        1       4
        2       5
        3       6
state # 3:
state # 4:
state # 5:
state # 6:
state # 3 accepts: [4]
state # 4 accepts: [2]
state # 5 accepts: [3]
state # 6 accepts: [1]

....

++++++++++++++++

Can anybody tell me what states

2 in NFA (dead state?)
3 in DFA (corresponding dead state resulting from subset construction?)
2 in DFA (alternative start state but still dead?)

are good for?

I understand that "accepts: [1]" indicates a final state accepting "v".
What is "accepts: [4]" trying to tell me? I can only see three rules...

Thanks for your help.

Uwe.