ICO contributors get both.
The ETH address you contributed from becomes one of 3,333 TimeMints (similar to Masternodes) that is able to create new DAY tokens through the passage of time.
On day zero, you will initially receive 24 DAY tokens per ETH. After day zero your TimeMint starts minting additional DAY tokens everyday.
In the whitepaper, the minimum needed to set up a TimeTX contract is 8,888 DAY. This means you need something like a 200 ETH investment to sell your TimeMint at the end of the first chrono era (if you just let the DAY generate more of itself) and 300+ to be able to sell it early on, even for the 1% mints. In practice, since the Mint presumably isn't mining anything while it's in the contract (as your DAY that you were mining with just got locked) it's even worse. There's also no real way to check whether anyone has a mint for sale in the first place (I guess people will wind up posting in here? meh) so it's extremely illiquid.
It's an interesting concept but I strongly recommend revising this point.
Additionally, to be blunt, the use cases in the white paper are anemic. The WP describes a token designed to be as hoarded as humanly possible, but there's nothing to spend it on and anyone without a Mint (remember, they're effectively not tradeable) has no reason to accept it. Investors are dependent on the company to eventually launch a platform that may, one day, do something with those tokens. "We'll find a lot of use cases later" is not good enough to send 88 ETH off and get back an asset it will take 300 more to ever trade.
The team's pedigree is beyond solid and the token itself is unique, so I'm puzzled why this ICO had to be now and not two weeks of use case brainstorming / development later, because there must be something better to shoot for than this.