Re: Millikan Apparatus- best one ?

[sample@lyon.edu (John D. Sample)]

At 12:57 PM on 10/10/96, <phys-l@mailer.uwf.edu> wrote:
> > Surely the ability to get a number for e is important in the Millikan oil
> > drop experiment! There are "easier" experiments for measuring e/m, h/e,
> > etc., but this is one where you are able to measure one of the fundamental
> > constants all by itself--you then obtain values for the others.
>
> That's amazing. How did you obtain drops with charge of exactly e?
> All my drops had charges which I determined to be *near* ne, where
> n is an unknown integer. The distribution must be subjected to some
> statistical procedure to infer the value of n for each drop. What
> procedure did you use to determine that the charge was quantized?
> How did you measure n?
>
> Leigh

I think I had the same experience as you.  As a student we measured only 5
drops, but one had very nearly twice the charge of another.  The other
three had approximately 3/4 the charge of the smaller of the first two, so
we guessed the last three each had three units of charge apiece, and then
one had four and the other 8.  Using least squares we calculated
e=1.675E(-19) C.  Unfortunately we weren't required to calculate the
uncertainty in this, only the percent error.

Even using so few drops there is little possibility that all of them might
have an even # of charges and so would yield 2e instead of e.

Chip