ELECTRODE EFFECTS 



565 



shape of the cur\e for phenol fibre is apparently due to a non-uniform- 

 ity in the samples which will be discussed later. 



That the abo\e results are materially affected b>- flux passing partly 

 through air and partly through the sample was proved by an addi- 

 tional test. The •^s" i^henol fibre sample with 4W' discs on each side 



k 



S 



'/ 02 pj ct 



THJCKN£:53 - //VC/Y£5 



Fig. 3 — Percentage edge correction for samples 6 inches square with foil electrodes 

 both 4^-2 inches in diameter 



was first measured in air. Since the upper electrode could not be 

 entirely covered and still obtain contact with it, the \i" phenol 

 fibre plate was laid on top of it so as to cover just one-half of the 

 sample and electrode. This caused an increase of 1.4 mmf. Using 

 the M" plate in the same way to cover one-half of the lower surface 

 of the sample and electrode, the increase was .7 mmf. showing that 



