The Added of Homomorphism in a Lattice Factor

With latisus, the trick seems to be that the mapping from your secret to the public value, they're both in the integers. And so obviously the added of homomorcism works. If you kind of multiply the secrets in the of appropriate way, you still get intagers out. That actually corresponds to a not completely crazy operation that you can do on the latice factors. So when you add to cipher tex, that corresponds to adding the two messages, but also adding the two noise terms. The noise gets a little bigger, ok, fine. You make room for it. But then when you multiply two cypher texts It's like making sure there's enough room at