PYROMETER 113 precise results with this method requires care- ful attention to several operations in the pro- cess, such as allowance for loss of heat by radiation from the water vessel during the ex- periment, and loss of weight of water by evap- oration after weighing it and after the immer- sion of the heated platinum. We have also to guard specially against the projection of water from the apparatus by the generation of steam by the hot platinum. 7. Becquerel's pyro- meter, based on the strength of thermo-electric currents produced by heating the junction of two different metals, is an improvement on a similar one devised by Pouillet. Two wires, one of platinum and the other of palladium, each about two metres long and of one square millimetre of section, are firmly tied together with fine platinum wire for a distance of about a centimetre from their ends. The palladium wire is enclosed in a porcelain tube, while the platinum wire is on the outside of this tube, which is itself enclosed in another tube of porce- lain. The free ends of the palladium and plati- num wires are soldered to copper wires which lead to a tangent galvanometer, and the junction of the copper and the palladium and platinum wires are immersed in melting ice to keep them at a constant temperature, so that no thermo- electric current can be generated in the appara- tus except at the junction of the wires in the porcelain tube. In order to obtain the value of a high temperature, the end of the porcelain tube containing the junction of the wires is heated up to this temperature, and from the deflection of the galvanometer needle produced by the thermo-electric current thus produced we deduce the temperature of the junction of the wires. This apparatus, to be of any value, has to be graduated by exposing the junction of the metals along with an air thermometer to the same successive high temperatures, and thus fixing the relation connecting the indica- tions of each apparatus with the correspond- ing temperatures. 8. Siemens's pyrometer is thus described by the inventor : " In order to realize a pyrometer by electrical resistance, it is necessary to rely upon the absolute mea- surement of the electrical resistance of a coil of wire, which must be made to resist intense heats without deteriorating through fusion or oxidation. Platinum is the only suitable metal for such an application, but even platinum wire deteriorates if exposed to the direct action of the flame of a furnace, and requires an exter- nal protection. The platinum wire used has, moreover, to be insulated and supported by a material which is not fused or rendered con- ductive at intense heats, and the disturbing in- fluence of the varying resistance of the wires leading to thje platinum wire has also to be neutralized. These various conditions are very fully realized by the arrangement represented in %. 4. Thin platinum wire is coiled upon a cylinder of hard-baked porcelain, upon the surface of which a double-threaded helical groove is formed for its reception, so as to prevent contact between the coils of wire. The porcelain cylinder is pierced twice longi- tudinally for the passage of two thick plati- num leading wires, which are connected to the thin spiral wire at the end. In the upper portion of the porcelain cylinder the two spi- ral wires are formed into a longitudinal loop, and are connected crossways by means of a platinum binding screw, which ad- mits of being moved up or down for the purpose of adjustment of the electri- cal resistance at the zero of the centigrade scale. The porcelain cylinder is pro- vided with projecting rims, which separate the spiral wire from the surrounding protecting tube of plati- num, which is joined to a longer tube of wrought iron, serving the purpose of a handle for moving the instrument. If the tem- perature to be measured do not exceed a moderate white heat, or say 1,300 C.=2,372 F., it suffices to make the lower protecting tube also of wrought iron to save expense. This low- er portion only, up to the conical enlargement or boss of iron, is exposed to the heat to be measured . Three leading wires of insulated copper united into a light cable connect the pyrome- ter with the measuring in- strument, which may be at a distance of some hundred yards f roni the same. They FIG. 4.; Siemens's Py- rometer, Coil Tube. are connected by means of binding screws at the end of the tube to three thick platinum wires passing down the tube to the spiral of thin pla- tinum wire. Here two of the leading wires are united, whereas the third traverses the spi- ral, and joins itself likewise to one of the two former, which forms the return wire for two electric circuits, the one comprising the spiral of thin wire, and the other returning imme- diately in front of the same, but traversing in its stead a comparison coil of constant resis- tance. By this arrangement of wires the ef- fect of the varying resistances of the leading wires is completely neutralized, for both bat- tery circuits comprise the leading wires up to the distant coil, and all variations of resistance by temperature to which the leading wires may be subjected affect both sides of the bal- ance equally. The measuring instrument may consist of a differential galvanometer if to the constant resistance a variable resistance be added. If the pyrometer coil were to be put
