Question about Mean-Variance Optimisation

Hi everyone, I was reading the MVO example in Curriculum (volume 1, page 296), which is shown below (I hope the picture is uploaded successfully).

In this graph, three different risk aversion coefficients (lambda) are depicted. However, my understanding was that, for a certain lambda, the line will intersect with the efficient frontier and the intersection is the optimal portfolio for this client. Why the lines are not intersecting?