map is your friend, oneliner F:\dev\oeufmayo>perl -MData::Dumper -e"print Dumper [ map{[split(':',$_)]} split(';',shift) ]" a1:a2;b1:b2;c1:c2; and, more than one line

#!/usr/bin/perl -w
use strict;
use Data::Dumper;
my $string = shift; # equivalent to shift @ARGV;
my @DD = (
           map
           {
             [
               split( ':' , $_ )
             ]
           }
           split( ';' , $string )
         );

print Dumper \@DD;
__END__
$>perl DD.pl a1:a2;b1:b2;c1:c2;11:22:33:44:55:;3:4:5;
$VAR1 = [
          [
            'a1',
            'a2'
          ],
          [
            'b1',
            'b2'
          ],
          [
            'c1',
            'c2'
          ],
          [
            '11',
            '22',
            '33',
            '44',
            '55'
          ],
          [
            '3',
            '4',
            '5'
          ]
        ];

$>
[download]

If any of the above isn't clear, please read perlfunc:map, perlfunc:split, perlman:perlref, and perlman:perldata.

Also, if you do not wish to use map, but instead continue with your for loop deal, you should take the time and read perlfunc:push, that way you do not have to keep an index of any of your arrays.

 
___crazyinsomniac_______________________________________
Disclaimer: Don't blame. It came from inside the void
perl -e "$q=$_;map({chr unpack qq;H*;,$_}split(q;;,q*H*));print;$q/$q;"