|Evolution of DNA -
Cell membranes were very cool, and they let Cassius live in all kinds of new places. But they also made life more complicated for Cassius. Now that it was enclosed in its own little bubble, it needed to do a more careful job of splitting its genetic material, when cells divided. If either of the baby cells missed out on any of the important genes, they'd die very quickly.
That meant that there was a huge selective advantage available, for any cell that could insure that both its daughters had a full set of genes, after each cell division.
One way to ensure that there was sufficient genetic material for both cells was to simply create many duplicates of all the genes, shortly before cell division. Most likely that was the earliest method that cells used, but it had a metabolic cost. Replicating many extra copies meant that the cell would divide more slowly, since it had to create all that extra RNA.
As cells became more complex, it also became dicier for cells to rely on random chance, for getting sufficient genes into each half during a division. The more genes there were, the more chance of failure. Anything that could increase the reliability of equal gene splitting would have offered a huge evolutionary advantage to early cells.
Another way to better manage cell replication would be to develop a 'smart' Nathaniel that would keep just one copy of each chain, and then replicate itself and one of each gene, when it came time for a cell division.
That's an easy process to describe, but not so easy to evolve! How would a Nathaniel 'know' how to connect to just one copy of each gene? How would it add new, useful genes? How would it really make the replication?
Unfortunately, it's probably too early to be looking at proteins clever enough for that type of gene management . So let's look at a different approach that is much easier to implement.
We've already talked about the use of operons and stop codons, as a way to put related genes close together. Combining genes also had a beneficial side effect, since it reduced the number of genetic chains, which in turn made it easier to perform successful cell divisions.
You could just carry that process further, and consolidate genes into a very small number of chains (called chromosomes). Then you'd only need one replication, and some moderately clever Nathaniel that made sure that each baby cell got one of each chromosome. That is much easier to arrange than any efforts to corral hundreds of small chains to the correct half of a dividing cell .
Of course, to locate each gene within a long chain, Cassius absolutely required an ID marker at the start of each gene, and stop codons at the end. However we have already talked about their appearance in earlier organisms, as a result of other selective pressures. Switching to a single genetic chain would have simply completed the adoption of an ID system for every gene.
Because successful cell division was so vital, it seems likely that Cassius would have reduced its chain count not long after the introduction of gene IDs and cell membranes.