Solution

ID: past-exam-of-the-mathematics-course-of-the-university-of-cambridge/2006/ii/paper-3/21g/solution

If and , then because is bounded and linear. Therefore , proving weak continuity of .
For a bounded sequence in , extract successive subsequences on which the first, second, and subsequent coordinates converge, then take a diagonal subsequence. Call the coordinate limits . Every finite partial sum satisfies , so and . For any , the pairing on its first coordinates converges, while the pairing on the tail is bounded uniformly by . First choose large, then take the subsequence limit. The Riesz representation theorem identifies these pairings with every continuous linear functional, proving the requested weak convergence.
For , norm convergence implies weak convergence immediately. Conversely, suppose but a subsequence has . Coordinate convergence to zero allows us to select indices and cutoffs such that the mass of before or at is below , and its mass after is below . Its mass on the intervening block is then at least . On that block set when nonzero and zero otherwise. These prescriptions define . The bounded linear functional has real part at least on every selected , contradicting weak convergence. Translating by the limit proves weak and norm convergence of sequences coincide in , the Schur property.
A compact operator sends bounded sets to relatively norm-compact sets. Every bounded sequence in has a weakly convergent subsequence by the argument above, after scaling into the unit ball. Its image under converges weakly and therefore in norm by the Schur property. Sequential compactness in a metric space gives relative compactness of the image of the unit ball. Every bounded linear operator from to is compact.

New to topics? Read the docs here!