« first day (2832 days earlier)      last day (418 days later) » 

02:48
@JoonasIlmavirta Correct!
The a is a vocalisation of n.
In renderings of earlier Greek, the n/a/o morpheme may be written as an n with a small a/o on top.
It could also be an/on/na/no, I believe, depending on period and dialect.
03:08
@Cerberus That sounds vaguely similar to how we got septem/hepta, with different coping strategies for -ptm.
@Cerberus It's interesting that early Greek orthography shows the origin so well. Does it go back to PIE?
03:31
@JoonasIlmavirta Ahh yes, quite possibly!
Was the Indo-European word something like septm?
@JoonasIlmavirta I don't know.
As you say, other linguistic branches seem to have different coping-mechanisms.
@JoonasIlmavirta Oh, that is not actually how they wrote it then!
It is how we now render the earlier Greek.
So it is a modern, artificial, academic notation in certain contexts.
To show that the n was vocalised.
 
5 hours later…
08:39
@Cerberus Oh, I see. That makes more sense.
@Cerberus I think so. That'd lead to the correct Latin and Greek.

« first day (2832 days earlier)      last day (418 days later) »