OF ARTS AND SCIENCES. 449 



The instrument, as thus constructed, was found to be extremely 

 sensitive, and admirably adapted to detect the smallest difference of 

 potential of any two bodies, but for a long and continued series of ob- 

 servations, or single experiments of long duration, it was found unser- 

 viceable. The electric field of the quadrants was not sufficiently 

 protected from external electrical influences. Two sides of the case 

 only were coated on their interior surfaces with tinfoil, but the the- 

 ory of the instrument demands, as far as possible, a complete shielding 

 from external electrical influences. The great delicacy of the sus- 

 pension is possibly the cause of the most serious defect, viz. a shift- 

 ing of the zero point. The needle would never return exactly to its 

 initial position. The difference was often great enough to give on a 

 scale distant a meter from the mirror, a deflection of a centimeter. 

 In the course of an hour a change in the position of the zero, when all 

 the quadrants were connected, of from two to five or more millimeters 

 would occur. These changes were in part due, no doubt, to a loss in 

 the charge of the needle. To remedy the first defect, new suspending 

 filtres were inserted without effect. To remedy the change due to 

 dissipation of the charge, the vessel jar was paraffined around the 

 edge, and for a time better results were obtained, though still faulty. 

 The glass rods supporting the quadrants were several times taken out, 

 washed with alcohol, dried, paraffined around the edges, and replaced. 

 Other parts of the instrument, in which it was thought the fault might 

 lie, were also carefully cleaned, and where needed provided with better 

 insulation. 



To obviate the difficulties met with in using the Clifton, the instru- 

 ment described in the following sections was designed by Professor 

 Trowbridge, and made at Boston, Mass. It is essentially the Clifton, 

 so modified as to retain its great sensitiveness without having the 

 defects mentioned. It was also desired to have an electrometer of 

 more convenient arrangement than any of tiie forms now in use, — one 

 in which the different parts should be amjily protected from external 

 influences, and yet be easily accessible for examination. The instru- 

 ment devised has two compartments. In the upper compartment are 

 the quadrants, needle, and suspending apparatus. In the lower com- 

 partment is the glass jar, with the arrangement for charging the 

 needle. The upper compartment consists of a wooden case, 25 cm. 

 high and 20 cm. square. On the top and back of the case are tight- 

 fitting brass doors, the one at the top being 12 cm. square, the door 

 at the back being 16 cm. high and 12 cm. wide. When open, these 

 doors allow easy access to the suspending frame, and when shut, form 

 VOL. XX. (n. s. XII.) 29 



