He had scarcely completed installing the thing, whatever it was, when a gentle tap at the door startled me.  Kennedy nodded, and I opened it.  It was Carton.

“I have had my men watching the Mayfair,” he announced.  “There seems to be a general feeling of alarm there, now.  They can’t even find Loraine Keith.  Brodie, apparently, has not shown up in his usual haunts since the episode of last night.”

“I wonder if the long arm of this vice trust could have reached out and gathered them in, too?” I asked.

“Quite likely,” replied Carton, absorbed in watching Kennedy.  “What’s this?”

A little bell had tinkled sharply, and a light had flashed up on the attachments to the apparatus.

“Nothing.  I was just testing it to see if it works.  It does, although the end which I installed down below was necessarily only a makeshift.  It is not this red light with the shrill bell that we are interested in.  It is the green light and the low-toned bell.  This is a thermopile.”

“And what is a thermopile?"’ queried Carton.

“For the sake of one who has forgotten his physics,” smiled Kennedy, “I may say this is only another illustration of how all science ultimately finds practical application.  You probably have forgotten that when two half-rings of dissimilar metals are joined together and one is suddenly heated or chilled, there is produced at the opposite connecting point a feeble current which will flow until the junctures are both at the same temperature.  You might call this a thermo-electric thermometer, or a telethermometer, or a microthermometer, or any of a dozen names.”

“Yes,” I agreed mechanically, only vaguely guessing at what he had in mind.

“The accurate measurement of temperature is still a problem of considerable difficulty,” he resumed, adjusting the thermometer.  “A heated mass can impart vibratory motion to the ether which fills space, and the wave-motions of ether are able to reproduce in other bodies motions similar to those by which they are caused.  At this end of the line I merely measure the electromotive force developed by the difference in temperature of two similar thermo-electric junctions, opposed.  We call those junctions in a thermopile ‘couples,’ and by getting the recording instruments sensitive enough, we can measure one one-thousandth of a degree.

“Becquerel was the first, I believe, to use this property.  But the machine which you see here was one recently invented for registering the temperature of sea water so as to detect the approach of an iceberg.  I saw no reason why it should not be used to measure heat as well as cold.

“You see, down there I placed the couples of the thermopile beneath the electric furnace on the table.  Here I have the mechanism, operated by the feeble current from the thermopile, opening and closing switches, and actuating bells and lights.  Then, too, I have the recording instrument.  The thing is fundamentally very simple and is based on well-known phenomena.  It is not uncertain and can be tested at any time, just as I did then, when I showed a slight fall in temperature.  Of course it is not the slight changes I am after, not the gradual but the sudden changes in temperature.”