With the MCIS cart removed, Geoff hooked up the leak detector to the MWS source gas line. He did a couple of hours of leak checking in this mode. Then he moved the connection to an empty gas port on the manifold itself. Much qicker responses. We replaced a VCR seal on the bottle side of the right-hand He solenoid valve. That did wonders. He checked around afterwards, but found no further leak. We tidied up and opened the manifold to IOS. Base pressure was about 1e-6 Torr and holding. After an hour or more, I decided to reinstall the leak detector on the manifold to assist pumping. I suspect air trapped on the front-side of the flow controllers. Opening them wide caused a vacuum spike, which recovered. It all seems to be working, and that is where we are right now.