The regex you provide to split specifies a separator, so Perl is splitting your string into chunks separated by sequences of ten characters each. What you get is the following:

$VAR1 = [
          <chunk1>,
          <separator>,
          <chunk2>,
          <separator>,
          <chunk3>,
          <separator>,
          <chunk4>,
          <separator>,
          <chunk5>
]
[download]

And all the chunks are empty except for the last one, which makes sense since there are no constraints on the separators (other than their length): Perl searches for the separator, finds it right at the beginning of the string, and splits it off along with a preceding empty chunk; then the whole process repeats, until you've only got five characters left, not enough to match the separator, so that's your last chunk.

Here's how I'd split a string into ten-character chunks using a regex:

#!/usr/bin/perl

use strict;
use warnings;
use Data::Dumper;

my $input = "1234567890abcdefghij0987654321ABCDEFGHIJlmnop";

my @chunks = ($input =~ /.{1,10}/g);

print Dumper \@chunks;
[download]

This produces:

$VAR1 = [
          '1234567890',
          'abcdefghij',
          '0987654321',
          'ABCDEFGHIJ',
          'lmnop'
        ];
[download]

Note that the quantifier needs to be {1,10} rather than just {10} here to accomodate the final chunk of less than ten characters. The regex engine's greediness will ensure that all chunks but the last will get the full ten characters.