Experiments on various Liquids. 155 



(2) "When the prime conductor gives no sensible spark in 

 the test for insulation, the first method fails, and the liquid is 

 then tried as a conductor, but first without the jar. The wire 

 from prime conductor to cell is broken, and the ends are in- 

 serted in two metallic balls, which are supported on insulating 



pillars, and separated by an air-interval. The machine being 

 set in motion, the two balls are put in contact at every turn 

 or half turn of the plate, so as to discharge the prime conductor 

 through cell to earth; and the polariscope is watched for resto- 

 ration at the instants of discharge. When the optical effect 

 is moderately strong and not very abrupt, the hand-compen- 

 sator works as well here as in the former case. There are 

 many liquids (nitrobenzol, benzonitrile, amyl alcohol, &c.) 

 which give intense restorations purely positive or purely nega- 

 tive when tried in this way, though they give no sure trace of 

 effect when tried as nonconductors. 



(3) When this method fails, the power is increased by the 

 addition of a Leyden jar, whose knob and outer coating are 

 connected respectively with prime conductor and earth. The 

 machine is set in motion, and the discharging balls are brought 

 into contact, first at every turn of the plate, then at every second 

 turn, and so forward, till the optical effect is brought out with 

 sufficient intensity. There are many liquids (butyl and propyl 

 alcohols, propionic and acetic acids, &c.) which give pure and 

 strong restorations at and above some point of this process, 

 but no trace of effect without the jar. 



(4) When these methods fail, the Winter's plate machine 

 is abandoned, and a Euhmkorff 's coil is applied. The coil is 

 worked by six large Grove's elements in series, and gives a 

 powerful spark (3 to 10 inches in air). The cell is sometimes 

 inserted in the circuit of the secondary coil by unbroken wires; 

 in other cases, there are air-intervals left at the outer ends of 

 the cell, which are traversed by spark discharges. It is found 

 to be generally advisable, and sometimes necessarv, to dis- 

 pense with the automatic rheotome, and to work the instrument 

 by hand, so as to obtain separate discharges at intervals of one 

 or more seconds. With all precautions, the discharges are 

 often too strong for perfect purity of electro-optic action, pro- 

 ducing large disturbances in the liquid, and giving rise in a 

 little time to intense heat-effects. There are several liquids 

 (ethyl alcohol, distilled water, <kc.) which have given good 

 results by this method only. 



2. The Fusion-cell, a small instrument similar to the plate 

 cell, but adapted to higher temperatures, is represented in the 

 adjacent diagram. 



The shaded piece is a square of plate glass, half an inch 

 2 



