IOWA ACADEMY OF SCIENCE 
277 
almost satisfactory, as may be seen from the following observations. It 
was necessary to determine the deflection from the first few swings of 
the needle. The period of the electrometer needle was about 10 sec. 
Pressure 90 gms. 30 gms. 
Zero 00 0 0 00 0 0 0 
D 1 ...+33 +33 —43 —43 —82 '—83 —80 —80 
D 2 . . . 1 1.5 — 2 — 1.5 — 3.5 — 3.5 — 3.6 — 4.0 after 10 sec. 
— 3.2 after 20 sec. 
— • 2.2 after 30 sec. 
Mean of D I 2 1.5 
Mean 3.65 
Corrected mean . . . .4.45 
Distance .00031 cm. .00043 cm. 
Since heating the mica and plates to 215° had made the insulation 
so nearly satisfactory, it seemed wise to repeat this treatment. The 
plates were kept covered with Japanese silk tissue to prevent dust par- 
ticles from settling on the plates. After six hours the plates were re- 
moved while hot and tested at once. The insulation was not perfect but 
it was improved and at the same time the capacity was larger. At. the 
beginning of the test the natural leak was such that the electrometer 
needle went from 3 div. to 1 div. in 50 sec. At the end of the observa- 
tions the natural leak was from 2.7 to 1 in 20 seconds, from 1.0 to .5 
in 30 sec., showing that the insulation deteriorated with time, possibly 
owing to the formation of a moisture film as the plates cooled. The 
natural leak was corrected for in calculating the distance. The following 
observations were obtained before the insulation deteriorated seriously: 
Zero 
0 
0 | 0 
0 
0 
0 
0 
0 
0 
y 
0 [ 
Mean 
— 
Dt 
D2 10 sec. 
9.0 pec. 
32 
.25 
32 
.5 
32 
.5 
-SI 
—1.0 
—SI 
—1.25 
—1.25 
—82 
—1.3 
—1.0 
—82 
—1.25 
—1.0 
—82 
—1.0 
—.85 
+83 
+1.0 
+83 
+ .8 
+ .4 
—82 
—1.25 
2 1.0 
1.1— 
.14 
Mean corrected 1.4. 
Capacity 1170. 
Distance, cm., .00013. 
I was not able by the microscopic method to determine the thickness 
of the mica flakes, further than to limit their probable size to less than 
five wave lengths. This much also was determined by a micrometer 
screw gauge. Then interference fringes were formed by bringing two 
optical flats on glass into contact along one edge and allowing the oppo- 
site edges to be separated by the mica flake. The number of fringes 
between the point of contact and the mica obviously gives twice the 
thickness of the mica in wave lengths. However, the great difficulty in 
this measurement was that sometimes considerable pressure had to be 
