PROCEEDINGS OF THE POLYTECHNIC ASSOCIATION. 419 



eacli otlier by levers, which are designed not to increase the motion, but to 

 communicate the sum of the motion of all the rods to a common wheel, 

 where the motion is increased by more delica'te mechanism. The sensi- 

 tiveness of this instrument is so great that the index fing-er starts up like 

 a thing of life upon the near approach of an}' person. Its size prevents its 

 use for any purpose except to note the changes in the temperature of the 

 surrounding air. 



The most delicate test of the presence of heat for a limited period is the 

 thermo-electric pile. The slight current of electricity generated when one 

 end only of the pile is heated is conveyed to the ordinary galvanometer, 

 and, as increasing or diminishing the heat increases or diminishes the 

 strength of the current, the index of the galvanometer becomes an index 

 of heat. Melloni has made many interesting experiments with this instru- 

 ment, which is said to be so delicate as to indicate the comparative bodily 

 heat of insects. 



Mr. Garvey. — The metallic thermometer, there are grounds to suppose, 

 can never be relied upon as an accurate indicator, as we have to depend 

 on experiments in each case for its correctness. It occurs to me that we 

 might get a metallic thermometer in which some of these difficulties might 

 be eliminated; for instance, let us take a small piece of wood, free from 

 moisture, and to it we attach a hand, which is placed on a dial, and another 

 arm or stem of chemically pure metal, such as copper deposited by elec- 

 tricity. Holding this hand, with such an arrangement, I conceive it pos- 

 sible to get a sufficient amount of sensitiveness to indicate very slight 

 changes of temperature. The metal being homogeneous throughout would 

 be free from the errors arising from the use of steel. 



Mr. Beaumont. — The most sensitive metallic thermometer in use is that 

 invented by Breguet, which consists of five plates of platinum and five 

 plates of steel, soldered together with silver and rolll'd out almost as fine 

 as paper; this strip is coiled into a spiral and suspended from an upright; 

 at the lower end an index is fastened which moves over a dial. 



Mr. Dibbin. — How is it with your thermometer as to keeping in tune, as 

 it were? That is, after being once set, will it always come back to the 

 same point for the same amount of heat for which it was graduated? 



Mr. Beaumont. — The first thermometers I made were five degrees out of 

 place in two weeks, and I was puzzled to know the cause. I set them 

 aside for six months and then set them again, and now' they are right. 

 I attribute this variation to inequalities in the steel, and it becoming 

 expanded too much, or at least bej^ond the point at which it was graduated, 

 the elasticity of the steel not being enough to bring it back. 



Heat Regulators. 



The Chairman. — Tliis subject of thermometers recalls to mind tlie methods 

 devised for regulating the heat of stoves. These all originated at Seneca 

 Falls, in this State. About the year 1842 or 3, there were four or live law- 

 yers in Seneca Falls whi) devoted a large portion of their time to practical 

 mechanics. Among them was John Jilaynard, afterwards a judge of the 

 supreme court. At one period he spent all his leisure time iu the machine 



