n. ELECTRICAL RHEOMETRY. 29 



5. A magnetic needle, to be introduced into the globe or circle, to explore the 

 force of the current in different places. 



All these instruments are represented together in PI. III., Figs. I. to V., and we 

 propose to pass them in review. 



I. Electromotor. Various electromotors or voltaic combinations, of quite recent 

 invention, are considered as constant, and indeed they are so for a short interval 

 of time, when the circuit is open and offers a great resistance to electricity. But 

 having had several times the opportunity of using batteries of very different 

 construction, I was taught by experience that none of them is perfectly constant, 

 and cannot he so, especially when the circle of conductors is short, and presents a 

 very small resistance. In an appendix to this memoir, an account is given of 

 the comparative experiments made with all the most celebrated combinations, in 

 order to ascertain their constancy. The result of these and other experiments 

 proved that, when the connecting wire is long and thin, a nearly perfect con- 

 stancy may be obtained during several hours, which is what is required for great 

 accuracy in our experiments. We have preferred Daniell's combination to any 

 other. The copper plate was a cylinder or rectangular case of that metal, holding 

 a saturated solution of sulphate of copper, the zinc being in a bladder with a 

 solution of chloride of sodium. For large plates, we used with great advantage, 

 instead of the bladder, a bag of sail-cloth covered with a thick paste of flour and 

 slaked lime, in equal proportions, which is applied to the bag when it is moistened, 

 but is fit for use when it has been dried. We have found such a preparation to 

 answer better for the separation of liquids than bladder itself, even when the 

 canvas is not very thick. Its resistance is very little increased by that preparation; 

 it resists a good deal less than an earthen cell, and after having been used once or 

 twice, the bag acts better than the first time. The reasons for preferring this 

 battery to the others were, 1st, its greater constancy, and the facility with which it 

 can be kept in the same state, when compared with other combinations, caused by 

 a certain quantity of sulphate kept in the corners of the element, as is seen in 

 Fig. 1. 2d. There is no production of gas, deleterious to the health of the operator 

 or injurious to the surrounding apparatus, as in Bunsen's and Grove's combinations. 

 3d. There are no fluctuations when the resistance of the circuit is suddenly changed, 

 interrupted, or restored ; but after all these alterations, when the resistance becomes 

 the same, the needle points always to the same degree. These irregularities are 

 chiefly remarkable in all those batteries where there is a large development of gas, 

 and are very inconvenient for a good success in the experiments. Perhaps the 

 different quantity of gas which is adherent to the plates, and is detached from them 

 or increased during the alternations of force, is the origin of these fluctuations. 

 The combination of platina and zinc with sulphuric acid, which is very constant 

 when not touched, was found the most irregular by changing the resistance of the 

 circuit, and was worse than that of copper and zinc. 



Some preparatory experiments for ascertaining the constancy of the current, 

 were made by introducing into the circuit all the resistances which were to be 

 used, and the conclusion was that within six or seven hours there was no sensible 



