EXPERIMENT STATION BULLETINS. 569 



ment of the bridge, in either direction, as much as a tenth of a scale 

 division should cause a deflection in the galvanometer scale. Under some 

 conditions the range may extend a number of scale divisions before de- 

 flection occurs. Thus only an aj)proximate value is obtained and the 

 reading is not precise. By using a condenser this error can be corrected 

 to a certain extent, but it is far better to avoid the use of condensers and 

 use resistances which have no capacity. 



Again capacity efl'ects increase with the frequency and soon produce 

 heating which will raise the true resistance value and thus introduce an 

 error. At a frequency of a thousand cycles per second the current does 

 not flow equally through the cross section of the conductor but is con- 

 fined mainly to its outer surface. Consequently when the current is not 

 distributed equally heating results and the resistance rises and there is a 

 false reading. 



By using alternating currents the errors from self inductance are 

 greatly increased. With direct current self inductance only occurs at 

 the making and breaking of the circuit. Alternating currents are not only 

 passing forward and backward, but are rising and falling in value and 

 such changes in a conductor will induce in it a counter e. ra. f. This 

 self-induced e. m. f. will be in such a direction as to oppose the change of 

 current which it produced. The amount of self inductance is dependent 

 on the rate of change in the current. Obviously alternating currents, 

 and these at a high frequency introduce a great error. These can all 

 be eliminated by using resistances wound non-inductively. The new 

 Curtis resistance coils are such, and they are the only suitable ones for 

 electrolytic measurements. 



RESISTANCE CELLS. 



The solutions or living plant or animal tissues to be studied, are placed 

 in certain resistance cups or cells. The form, shape and size of both the 

 cups and the electrodes are matters of importance. The accuracy of a 

 measurement is dependent to a great extent upon the design of the cell. 

 Freedom from impurities must be assured. The construction of the vari- 

 ous parts must not be lightly considered. There are a number of factors, 

 which make it necessary to construct resistance cells of various forms. 

 Tn regard to solutions, one particular form is not suitable for concen- 

 trated and dilute solutions alike. The form of the cell is dependent upon 

 the size and shape of the electrodes and upon the distance these are apart. 

 In turn the nature of the solutions controls the size and shape of the 

 electrodes and the distance they should be apart. For convenience these 

 cells have been grouped according to the purpose they serve. 1. Resist- 

 ance cups for good conducting fluids. 2. Resistance cups for poor con- 

 ducting fluids. Several different forms are figured in the publication of 

 Kohlrausch and Holboru (14) and recently new forms by Prof. Washburn 

 are i>ictured in catalogue No. 48 of the Leeds & Northrup Company. The 

 Fritz Kohler catalogue for 1900 also shows a number of forms. 



The electrodes are usually circular in form and consequently the re- 

 sistance desired is the resistance of a cylinder of liquid enclosed between 

 them. This cylinder is modified greatly by either varying the distance 

 between the electrodes or increasing the surface of the electrodes. As a 

 general rule, it may be said that the electrodes must have a large surface 

 and be close together when the solution to be measured is "conductivity- 



