In a late communication by Mr. Smirke to the Institute of British Architects, he directed attention to cisterns for containing water, describing three different kinds as used in England. First, the hogshead sunk in the ground ; second, a wooden cistern lined with sheet lead ; third, the slate cistern, a vast improvement upon its predecessors, but too expensive for general adoption among the working classes. The object of Mr. Smirke was to bring to notice another material for cisterns, namely coarse earthenware of large size and low price. He had seen and measured earthenware cylinders of two feet five inches internal diameter, forming the lining of a well more than two thousand years old, amidst the ruins of ancient Selinuntum, and they were as sound as when first made. A vessel of this material, three feet square and four feet deep, would hold upwards of two hundred gallons. Coarse earthenware would no doubt be an excellent material for cisterns, but more expensive than brick and ? hydraulic cement, of which the majority of cisterns in this part of America are made. It would seem, from Mr. Smirke's communication that such cisterns are unknown to him. We would recommend them to the attention of the Institute of British Architects. They are made by excavating a deep circular hole in the ground, below the reach of frost, then laying a stratum of clay or cement on the bottom, paving the floor, and raising the sides with brick, surmounting the whole with an arched roof. The outside of the side walls should be coated with hydraulic cement, or puddled round between the ground with clay, and the whole interior plastered with a thick coat of hydraulic cement. A small opening is left for the entrance water pipe, and another for the suction tube of a pump. A square mouth covered with a lid is usually made at the center of the top for admission to clean it out, or for repairs. Such cisterns, if carefully constructed, will last for centuries.
This article was originally published with the title "Cisterns—Hint to Potters" in Scientific American 13, 33, 261 (April 1858)
doi:10.1038/scientificamerican04241858-261c