II. ELECTRICAL RHEOMETRY. 31 



The galvanometers were always kept distant from each other, and were supported 

 on brackets steadily fixed to a wall or on the sill of a window; care was taken to 

 remove all iron from their neighborhood, but no sensible influence was discovered 

 on the results, arising from this cause, on account of the system adopted of observing 

 constant deviations ; only strong magnets had a little influence, Avhich of course 

 were removed to a great distance. 



III. Rheostat. This instrument is already known to philosophers, and is 

 described by "Wheatstone, its inventor, in the Archives de TElectricite} It is repre- 

 sented obliquely in Fig. III., and in erect position, that it may be seen better, in 

 Fig. III. bis. It consists of two cylinders movable on their metallic axes; the 

 cylinders are each a foot long and supported by a rectangular frame, on which they 

 can be turned by means of two little cranks, a, h. The larger cylinder is about 

 two inches in diameter, made of hard wood, and varnished. The smaller one is 

 about an inch in diameter, and of copper. The wooden cylinder has a screw of a 

 very fine thread from one end to the other ; the copper cylinder is perfectly smooth 

 and polished. A thin copper or brass wire attached with one end to the top of the 

 wooden cylinder, and with the other to the end of the copper cylinder, unites them 

 both, and can be coiled upon one or upon the other. When the apparatus is intro- 

 duced into the circuit, it is evident that the part of wire coiled on the copper 

 cylinder aflfords no resistance at all to the current, and it is as if it were not in the 

 circuit. A divided scale, mn, is used to reckon the turns, and there is a circular 

 division at the end of the wooden cyhnder in o for enumerating fractions of turns 

 by an index i, attached to the frame. "Wheatstone used to put both cylinders in 

 communication with the current through the Y's which support the pivots, but by 

 working long, heterogeneous matters gather there, and the metal becomes soon 

 oxidized by electric action, and the current is often interrupted. I preferred, there- 

 fore, to terminate both upper pivots by two wheels an inch and a half in diameter ; 

 these are immersed into two small troughs, filled with quicksilver, thus allowing 

 a perfectly free passage to the current without sensible resistance or accumulation 

 of heterogeneous obstacles. 



It is indifierent whether the diameters of the cylinders be equal or not ; but if 

 the metallic one be smaller, the wire is coiled more tightly around the wood, which 

 is very useful in oi'der to have always the same length for the same number of turns. 



Interposing this instrument in the circuit, and regulating the length of wire coiled 

 on the wooden cylinder, the intensity of the current may be reduced to any degree 

 required. If the length of that wire be not sufficient, several other coils may be 

 introduced, whose resistance must have been previously determined. 



To facilitate the introduction into the circuit of such coils as are seen in B, R', 

 Fig. III., a small parallelopiped p, p' was used, in which were several holes or 

 cups containing quicksilver; the ends of the wires were immersed there, and a 

 little bridge B made of a thick copper lamina or wire was used to connect difierent 



1 Arch, de I'Electr., t. iv. No. 13, 1844, p. 102. Description de plusieurs instruments h, proc6d6s 

 nouTeaux pour determiner les constantes d'un circuit voltaique. See also Philosoph. Magazine. 



