138 REPORT — 1844. 



From this 4th set of experiments, it would seem that two warmed insula- 

 tors retained the charge as well as one ; and that therefore in the same situ- 

 ation each insulated '■'^ perfectly " using this expression in Coulomb's sense. 

 But circumstances may arise in applying this kind of test to render the con- 

 clusion defective. The two cajjs, &c. (having greater capacity and quantity) 

 should retain the charge better than one, &c. 



It is evident that certain mysterious conditions of the ambient air interfere 

 sometimes with our operations of this nature, and (as Sir David Brewster 

 justly remarks) experiments should be undertaken to find them out*. 



5. Experiments on Insulation by means of Chloride of Calcium. 



The object of these experiments was to ascertain how far it might be prac- 

 ticable to construct electrometers which should lose the lower and more usual 

 charges, received from the principal conductor, at given periods, in some near 

 approximation to constant rates, yet not lower the tension of the conductor 

 materially, on contact with it. For it is necessary to a more exact prosecu- 

 tion of our inquiries, that the true electrical state of the conductor, as regards 

 both tension and kind, should be known at certain intervals of the night more 

 accurately than it can be by means of the resinous electrograph described 

 at p. 126' 



In order to procure the greatest possible constancy of loss, it was (ob- 

 viously) very desirable to obtain the greatest possible retention, and for this 

 purpose non-conducting or semi-conducting laminae, coated on both faces 

 with good (or better) conductors, naturally presented themselves as being 

 capable of retaining low charges for very great lengths of time. But these 

 require, proportionably, much larger doses of electricity to produce equal 

 effects on tension electrometers than simple conductors (not thus " compen- 

 sated"), and would, consequently, lower the tension of the principal conductor 

 materially at the time of I'eceiving their charge from it. 



In endeavouring to discover the best means of retaining a small quantity 

 of tension (" frictional") (y) electricity a long time, I first employed receivers, 

 air-tight and of various dimensions, containing vertical rods of glass (cut from 

 the same piece) about 3| inches long and a ^ of an inch in diameter. They 

 carried horizontal brass wires from which were suspended pairs of natural 

 voltaic straws (as in fig. 4), and were coated with the best engravers hard 

 sealing-wax, applied by heating the glass sufficiently to melt the wax (not by 

 spirit varnish, which is far less effective). The receivers also contained 

 each 2 or 3 ounces of chloride of calcium (below). The electrometers were 

 charged (by an electrophorus) after lifting the receiver up from the flat 

 glass plate on which they were placed (with a little oil in the joint). 



These experiments proved that an electrometer originally charged to about 

 one inch divergence of Volta's No. 1, or standard pair, would retain for the 

 space of from 114< to 124 hours (by the above means), some remainder of its 

 charge t; also that small receivers were better than large, &c. 



But it soon appeared that after the straws had been for two or three days 

 exposed to the action of the chloride, tlvey became insulating in a very incon- 

 venient degree, for when the wires supporting them were touched (continu- 



* Vide Encyclopaedia Britannica, vol. viii. p. 589, seventh ed. 



1st. Wliat relation has the actual quantity of " dry steam," in a given measure of air, to 

 the insulating power of that atmosphere ? 



2nd. What relation has the temperature of such an atmosphere with its insulating power ? 



3rd. In what degree is insulation influenced by the density of the atmosphere ? 



4th. Has oxygen gas and dry steam a different insulating power from nitrogen, &c. ? 



The solution of this query would not serve our purposes perhaps. 



t An uncoated rod retained some remainder for 102 hours. Had the receivers been j)er- 

 fectly air-tight perhaps this would have insulated as well as the others. 



