I took out the tubes on one side of one section, ten in all, five over the furnace and five behind the bridge wall, and planed them in two longitudinally, and had the following revelation: The tubes over the furnace were entirely empty. In those back of the bridge wall a deposit of sediment appeared, only about an inch deep in the first one, and increasing regularly to a depth of 18 inches in the last one, which was not the tube receiving the feed-water. So the water fed into the last tube of each section deposited its sediment most largely in the first tube it reached, in which the circulation was least active, and had deposited it all before reaching the tubes over the furnace. The remaining long tubes were then cleaned, the tubes cut in two were replaced by new ones, and the boiler delivered to Mr. Allen. The next stage in its history was very funny. The purchaser, to save the cost of Croton water, fed his boiler from the Harlem River, and within a month it was found to be filled solid with salt. What was done about it I never heard. I thought I could sell the boilers where, as in New York City, they could be fed with water free from lime, and I made a few such sales, but the inspiration which led me to employ the second drum for superheating the steam had deserted me. I came to the conclusion that by making the first drum a large one, and not extending the nipples into the drum to trap a puddle of water, as I had done, I could superheat the steam in one drum. That was a blunder. I had underestimated the furious circulation, which carried a large amount of spray into the drum. I was misled by the quiet position of the water-level, as always shown in the glass gauge. Instead of superheated steam, I found the boiler to give very wet steam. That fault, of course, I could have remedied by returning to my first design. But I was discouraged by other things. The first, of course, was the impossibility of removing scale by any mechanical means. The most serious discouragement was a cracked header. The inclined tubes, on any plan for their use that I could then design, made cast-iron headers necessary. I had taken great pains to obtain perfect castings, making them of the best iron in baked molds in iron flasks, of uniform thickness, ⁵⁄₈ in., and ³⁄₄ in. where threaded, with cores held perfectly central and remarkably well vented, and felt that I could rely on their soundness; but this defect showed that I could not. So reluctantly I abandoned the manufacture of the boiler.