It might be easiest to just give an example. The Sanger Institute has sequenced the genomes of a bunch of different mouse strains. Those genomes won't typically contain chromosomes, but rather 100-200 thousand contigs and scaffolds. Why? Because you're rarely able to assemble entire mammalian chromosomes from NGS data (due to repeat regions, unequal coverage, etc.). Joining these contigs/scaffolds together to make chromosomes is a somewhat long and drawn out process that requires a lot of effort (the human and mouse genomes aren't even completely finished yet...though they're good enough for most purposes).