Another way is to employ a "Rectifier." This is a conductor which has the peculiar property of allowing current to pass one way but not the other. It thus eliminates every alternate current and changes the alternating current into a series of intermittent currents all in the same direction. Rectified current is thus hardly described by the term continuous, but still it is "continuous current" in the sense that the flow is always in the same direction, and so it can be measured by the ordinary continuous current instruments. The difficulty about it is that there is some doubt as to the relation between the quantity of rectified current which the galvanometer registers and the quantity of alternating current, which after all is the quantity which is really to be measured. How the rectification is accomplished will be referred to again in the chapter on Wireless Telegraphy. But to return to the thermo-galvanometers, as those are termed which ascertain the strength of a current by the heat which it produces, the simple little contrivance of Sir William Snow Harris has more elaborate successors, of which perhaps the most interesting are those associated with the name of Mr W. Duddell, who has made the subject largely his own. Besides their interest as wonderfully delicate measuring instruments, these have an added interest, since they introduce us to another strange phenomenon in electricity. We have just noted the fact that electricity causes heat. Now we shall see the exact opposite, in which heat produces electrical pressure and current. And the feature of Mr Duddell's instruments is the way in which these two things are combined. By a roundabout but very effective way he rectifies the current to be measured, for he first converts some of the alternating current into heat and then converts that heat into continuous current. If two pieces of dissimilar metals be connected together by their ends, so as to form a circuit, and one of the joints be heated, an electrical pressure will be generated which will cause a current to flow round the circuit. The direction in which it will flow will depend upon the metals employed. The amount of the pressure will also depend upon the metals used, combined with the temperature of the junctions. With any given pair of metals, however, the force, and therefore the volume of current, will vary as the temperature. Really it will be the difference in temperature between the hot junction and the cold junction, but if we so arrange things that the cold junction shall always remain about the same, the current which flows will vary as the temperature of the hot one. The volume of that current will therefore be a measure of the temperature. Such an arrangement is known as a thermo-couple, and is becoming of great use in many manufacturing processes as a means of measuring temperatures.