Without trying to figure out what you are actually trying to achieve with the file manipulation, here at least is a sample the does what you seem to want for the line manipulations:

use strict;
use warnings;

my @lines = map {[split]} <DATA>;

for my $line (@lines) {
    for my $wordIndex (0 .. $#$line) {
        substr $line->[$wordIndex], 0, 0, '$';
        print join (' ', @$line), "\n";
        substr $line->[$wordIndex], 0, 1, '';
    }
    
    print "\n";
}

__DATA__
one fish,
two fish,
red fish, blue fish
[download]

Prints:

$one fish,
one $fish,

$two fish,
two $fish,

$red fish, blue fish
red $fish, blue fish
red fish, $blue fish
red fish, blue $fish
[download]

Note in particular the use of strictures (use strict; use warnings;) and the use of Perlish for loops rather than C for loops.

Note too that for sample code it is useful if the code can be run with sample data provided - a __DATA__ section helps a lot with that.