One of the earliest forms of room heating using hot water was the use of long runs of iron pipe, identical to that used described above for heating horticultural structures. A significant change was heralded when instead of using these long lengths, short lengths were connected together horizontally using a series of 180-return return bends. This new arrangement formed relatively short, typically 2ft. 6ft. long, taller structures containing any number of tubes. The objective was to provide as large a radiating surface area in as small a space as possible. Numerous designs were subsequently developed; the main differences were how the pipes were connected together. Initially, double bend siphon connections were used, identical to those used in normal low-pressure heating systems. These were later replaced using box ends whereby the pipes were caulked. Two forms of box-ended radiators developed, one with the internal sockets and the other with external sockets. Later, rubber jointed coil radiators were developed, whereby the joints were made by compressing rubber rings between the faced ends of pipes. Typically, 3-inch or 4-inch diameter pipes were used in hot-water systems and smaller diameter pipes used for steam heating. Because of their inherent unsightliness, it was not very long before they were being hidden beneath decorative cases, often with mesh or some other mechanism to allow heat to escape.
Thomas Messenger embraced the use heating coils, although they were not installed in great numbers. Between 1867 and the end of 1874, he was involved in installing or quoting for heating coils on forty-five occasions, with the vast majority being in a residential situation either for conservatories or the residence itself.
Unsurprisingly the firm made their own coils and cases and the first recorded instance of the latter was in 1867, when he charged Revd. Charles Evans of Blackwall, near Wirksworth, Derbyshire £1 for the wood and labour to make a case and 7s. to paint it; although if the case was not used Revd. Evans was to be made an allowance of £3 13s.
By May 1868, the firm were producing cases with marble tops for they gave Mr. August Clarke, a grocer of North Street, Loughborough an estimate for heating his house, which included one or more coil cases. The estimate included £1 5s. for the wood and labour, and £2 10s. for 10 superficial feet of marble.
The first indication of sizing was in July 1868, when Messrs Haughton & Thompson, agricultural implement manufacturers, of Botchergate, Carlisle, Cumberland, ordered four coil cases, one 3ft. 6in. long (£4); one 4ft. long (£4 4s.); one 4ft. 6in. long (£4 8s.); one 5ft. long (£4 18s.).
One of the earliest records that exist which describe the type of the coils that Thomas Messenger was marketing was at the end of June 1869. He delivered three estimates to Mr. G.T. Dodson of Kibworth House, Kibworth Harcourt, Leicestershire; the first estimate for heating the entrance hall included two 4ft. 6in. long coils consisting of 26 tubes, costing £6 10s. each; the third estimate was for providing heating under the stairs which included another 4ft. 6in. long coil, costing £13, to be comprised of joining two separate coils together.
A month later, the firm installed two coils in the hall at Maplewell Hall, near Woodhouse Eaves, Leicestershire for Charles Ashton. Both coils consisted of 26 tubes, one 4ft. long and the other 5ft. In addition, one coil had an elaborate coil case in oak, with lacquer gauze and a marble top, to match the fireplace. The 4ft. coil cost £5 10s., the 5ft. cost £6 and the case £7 10s.
The cost of hiding ordinary coils could be an expensive business. In early 1873, the firm moved a coil for Mr. John Stafford, a cheese factor and cigar merchant, of Elmsleigh, Stoneygate, Leicester. The price for repositioning a coil, including 7 days of a fitter and labourer’s time, amounted to a little over £10. Whilst the new coil, which was no doubt very decorate and similar to an existing one in the hall, was an apparent excessive £28 15s.
Another elaborate coil case was installed for a Miss Child, in 1873. The firm quoted for three coils, although only two were installed. The first was 2ft. long, employing six 3-inch tubes and installed in the pantry. By this time, the firm had started using india rubber joints for connecting the tube ends. This method led to a coil design that had both neater and more compact ends compared with the normal method of connecting the tubes with siphons and cement joints. The second, a double coil for the hall, was of normal construction, 2ft. 6in. long, composed of fifty-four 2-inch tubes. The accompanying case was made of oak with what was described as “Derbyshire slab” and cost £5 7s. 6d. As would be expected, the relative cost of the new coil design appears to have been significantly higher size for size compared with that of the normal method of construction. The price of the 3-inch tubes was 3s. 9d. per yard for the new coil compared with only 1s. 5d. per yard for the normal coil. The siphons were also more expensive at 4s. compared with 1s. 9d. In addition, the india rubber joints were 8d. each compared with 3d. for standard cement joints. The additional cost of the new design coil may have been the reason for using the normal design for the larger double coil, particularly as it could be covered with a decorative case at a combined cost that was probably less than that for an exposed india rubber jointed coil.