G'day Scotmonk,

Here's some basic logic which may achieve what you want.

#!/usr/bin/env perl

use strict;
use warnings;

my $COUNT = $ARGV[0];
my %seen;
my @data;

unshift @data, split while <DATA>;
@data = grep !$seen{$_}++, @data;
print "@data[0 .. $COUNT - 1]\n";

__DATA__
1 1 2 2 3 3 4
5 6 7 8
5 6 6 7 8
5 6 7 8
9 10 11 12
13 14 15 16
17 18 19 20
17 18 19 20
21 22 23 24
[download]

Some sample runs:

$ ./pm_11109149_reverse_file_extract_with_conds.pl 24
21 22 23 24 17 18 19 20 13 14 15 16 9 10 11 12 5 6 7 8 1 2 3 4

$ ./pm_11109149_reverse_file_extract_with_conds.pl 12
21 22 23 24 17 18 19 20 13 14 15 16

$ ./pm_11109149_reverse_file_extract_with_conds.pl 7
21 22 23 24 17 18 19
[download]

Notes:

• This solution is intended for small files (such as the sample input you show). Some of the techniques shown may also be valid for larger files — such as using %seen for duplicate checking — but you probably won't want to read millions of numbers into an array if you only want seven.
• I've modified your data to include duplicates in various places. For future reference, it's useful for us if you provide sample input with examples of this sort of thing.
• I've just used the DATA filehandle; you'll need to add your own I/O handling.
• I haven't included any exception handling; you'll need to do this. Areas to look at would be I/O errors, $COUNT not being an integer, and so on.

"I am just setting out learning PERL, ..."

That's fine: none of us left the womb knowing any Perl.

[Aside: PERL is wrong: the language is Perl and the executable is perl (which may have an extension on some systems, e.g. perl.exe on MSWin).]

If there's any code that I've used that you don't understand, try the Online Perl Documentation in the first instance — if still in doubt, ask another question here.