In previous correspondence Donaldson had asserted that the front row of roof stays in the Thornycroft boiler were arranged so as to allow the tube plate to expand. J. A. questioned this practice. He suggested that if they were inserted on the assumption that the expansion of the tube plate raised the crown, and with it the front row of roof stays, and that if they had to be fixed in some way so as to prevent a thrust being thrown on the stay, then they could be dispensed with altogether.