Some Experiments on the Latent Heat of Steam. 41 



With the Berthelot apparatus and the modification of 

 the same referred to above, the source of heat for the 

 boiler must lie underneath it and directly above the 

 calorimeter. Many different kinds of lids were used to 

 protect the calorimeter from the heat radiated by this 

 flame, but the total correction could seldom be reduced 

 below 3 % of the observed rise, and on this there remained 

 some little uncertainty. 



A metal apparatus was now designed in which the 

 heat was supplied by an electric current passing through 

 a coil of wire immersed in the water of the generator. 

 This is shown in Fig. 1. A cylinder of copper, A B, 

 8 cm. diam. and 20 cm. long, is provided above and 

 below with brass plates, which screw up against leather 

 washers, making a steam-tight joint. Through the lower 

 plate is fixed, centrally, a vertical brass tube 6 mm. 

 internal diam., which passes up the cylinder to just under 

 the upper plate and downwards to the steam-trap D, 

 consisting of a short piece of wider tubing having a 

 screw plug at F. 



From this trap the steam passes along the inclined 

 tube to the three-way tap C, by which it can either be 

 led away by the horizontal tube or downwards to the 

 condenser. The whole of the tubes outside the boiler 

 were wrapped with ^in. lead . pipe, known in the trade 

 as " twigging," through which a current of steam from a 

 second boiler could be passed. I found this a most 

 effectual and convenient way of jacketing any pipes which 

 had to be protected from radiation, and it was often used 

 in these experiments. 



The construction of the coil for supplying heat to the 

 water in the boiler gave considerable trouble. Although a 

 coil of bare wire could be wound on smooth supports so 

 tight as to be perfectly satisfactory when used in cold 

 water, yet at ioo° it always became loose, and short 



