Mr. Richards’ bed, shown in the illustration facing page 70, was designed in the box form, the superior rigidity of which had been established by Mr. Whitworth. It was a box closed at the top and flanged internally at the bottom. It rested on the foundation through its entire length. The main pillow-block was formed in the bed, as were also the lower guide-bars. The cylinder was secured on its surface in the old-fashioned way. [Illustration: Engine Bed Designed by Mr. Porter. Engraving made from an Old Print.] It occurred to me that the best features of the Corliss and the Richards designs might be combined to advantage. This idea I worked out in the bed shown in the accompanying illustration, taken from a circular issued by Ormerod, Grierson & Co., of Manchester, and which was made from a photograph of an engine sent by that firm to the Oporto International Exhibition in 1865. It will be seen that this is Mr. Richards’ bed with the cylinder bolted to the end after Mr. Corliss’ plan. The great strength of the bed enabled the supports under the cylinder to be dispensed with. This left the cylinder free to expand by heat, and made it convenient to attach the steam or exhaust connections or both underneath. This bed has remained without change, except in one important respect. I made the first cylinders with a bracket which was keyed up from the base of the bed. In the illustration a corner of this bracket appears. At the Paris Exposition in 1867 Mr. Beyer, of the firm of Beyer & Peacock, the Manchester locomotive-builders, when he saw it, told me I did not need that bracket. I then left it off, but found the cylinder to wink a little on every stroke when the heavy piston was at the back end. To find the weak place, I tried the following experiment on an engine built for the India Mills in Manchester. I filed two notches in the edges of the brackets on the bed, opposite each other and about ten inches forward of the head, and fitted a piece of wire between them. This wire buckled very decidedly on every revolution of the engine, when the piston was at the back end of its stroke. I then united these brackets into a hood, and lengthened the connection with the surface of the bed, as it is now made. This affords a perfect support for the cylinder. Experiments tried at the Cambria Iron Works on a cylinder of 40-inch bore and 48-inch stroke, with a piston weighing 3600 pounds and running at 100 double strokes per minute, showed the back end of the cylinder standing absolutely motionless. This experiment will be described hereafter. [Illustration: Cross-head Designed by Mr. Porter.]