# Why is $H^n(I \times Y, R) \to H^n( \partial I \times Y, R)$ a split injection?

In Hatcher's Algebraic Topology, section 3.2, during the computation of the cohomology ring of a $n$-torus, the following assertion is made. Let $Y$ be a space and $R$ a commutative ring. Then the natural map in cohomology

$$H^n(I \times Y, R) \to H^n( \partial I \times Y, R)$$

is a split injection. This is likely straightforward, and I believe it follows directly from the Kunneth theorem. However, I don't see a direct proof, which is probably what Hatcher is indicated. Can someone explain to me what I'm missing?

-

Isn't $I$ just the closed unit interval? Then $I\times Y$ is homotopy equivalent to $Y$ and $\partial I\times Y$ is just two copies of $Y$. It's clear that your map then is just the diagonal map from $H^n(Y)$ to $H^n(Y)\oplus H^n(Y)$.