D O C U M E N T 2 4 7 D E C E M B E R 1 9 2 0 3 4 3
As was to be expected, the sphere with 3 or more magnets (of 500 periods) is at
an adequate distance and centered; but, unfortunately, the iron distribution within
the sphere is asymmetrical (3 gyroscopes); hence preferred positions for the sphere,
since the lines of force penetrate through the sphere’s 3 aluminium
wall.[2]
An iron-sheet shield within the sphere still proved not to be sufficient. Even
a larger number of magnets (more than 10 cannot be comfortably housed, probably
for space reasons) will surely improve the situation but not solve it. There is simply
an asymmetry left.
Building upon your suggestion, I have now suggested in Kiel to try a ring mag-
net of about 120 diameter at a sphere size of 220 Thus something like
this:[3]
According to it, the construction
would be independent of the iron distri-
bution within the sphere. I do suspect,
though, that the efficiency of this ar-
rangement will be very bad.
Nevertheless, your suggestion will be
implemented in some way or other; I do
not doubt it and I would have liked to talk about this with you, acting on the thought
that, in these difficult times today, you might find amenable a sort of royalty in ad-
vance (for it will surely take years before a gyro-compass model can be put on the
market). Do you agree if the firm Anschütz & Co. delivered to you, for now, the
amount of 20,000 marks? I explicitly write deliver and not remit, because then there
would just be inquiries of a tax nature, of course. I would like to perform this de-
livery function sometime during your presence here, but if it simply does not work,
then there are other ways of having the amount reach you as well. Once the model
is for sale, then the co., or better yet, I myself shall settle accounts with you prop-
erly. Is it all right with you like this?
The heat-rotation experiment did not yield anything upon application of friction-
al
heat.[4]
Now I have set up a new experimental arrangement and send very hot oil
(approx. 200°) through a thick-walled copper cylinder. You will immediately un-
derstand the enclosed diagram. I suspect that with this arrangement disturbances
will occur less easily. I do still have some reservations about the use of a magnetic
needle to measure the ev. magnetism, since at closer proximities of a magnet to the
rotating copper cylinder, eddy currents must surely form, which might have a dis-
turbing influence on the experiment. Finally, there are other control measurements
as well that one can use once a positive result is in.
It goes without saying that I shall report to you further about the relevant
experiments.
My wife and I send you our warmest wishes for the year 1921; in particular, my
wife sends word that your agreement to come to Kiel in the summertime has been
mm-thick
mm mm.
sphere
ring
magnet
coil.