Answer: \(\boxed{r=0.654}\), and there is evidence of non-zero correlation at the \(5\%\) level.

(i) For the two regression lines, the product of the two regression coefficients is \(r^2\).

The gradients are \(0.46\) and \(0.93\), so

\[r^2=0.46(0.93).\]

Both gradients are positive, so \(r\) is positive:

\[r=\sqrt{0.46(0.93)}=0.654\quad\text{to 3 significant figures}.\]

Thus

\[\boxed{r=0.654}.\]

(ii) We test

\[H_0:\rho=0,\qquad H_1:\rho\ne0.\]

For \(n=12\), the two-tailed \(5\%\) critical value is

\[0.576.\]

Since

\[|r|=0.654>0.576,\]

we reject \(H_0\).

Therefore there is evidence, at the \(5\%\) significance level, of non-zero correlation between the variables.