Potier reduced Michelson's prediction by half. His mistake was to assume c+v to be affected by the mirrors. Instead mirrors control only c! Or? If we eliminate this mistake we can use the Galilean transform without time dilation. See attachments.
Regards from
John-Erik