The opinion of the court was delivered by: Juergens, District Judge.
This litigation was commenced by the plaintiff Armour Research
Foundation of Illinois Institute of Technology (hereinafter
referred to as Armour) against C.K. Williams & Co., Inc.
(hereinafter referred to as Williams), defendant in Civil Action
No. 3200 and plaintiff Armour against Technical Tape Corporation
(hereinafter referred to as Technical Tape), defendant in Civil
Action No. 3438. Minnesota Mining and Manufacturing Company
(hereinafter referred to as Minnesota) was subsequently, by order
of this Court, brought in as a necessary party plaintiff. The two
suits were consolidated for trial.
The plaintiffs charge infringement by the defendants of United
States Letters Patent No. 2,694,656, issued to Marvin Camras, who
filed his application with the United States Patent Office on
July 25, 1947. The patent issued November 16, 1954. The inventor
Marvin Camras assigned the patent to Armour. Minnesota is an
Armour licensee and has the exclusive right to grant sub-licenses
under the patent.
The defendants filed their separate answers to the respective
complaints, alleging invalidity and unenforceability of the
patent and denying infringement. In addition, Williams filed its
counterclaim seeking an adjudication of the invalidity and
non-infringement of the patent in suit and counterclaimed against
both plaintiffs for violation of the Anti-Trust Law.
This Court has jurisdiction under the patent laws of the United
In support of their allegations of infringement the plaintiffs
rely on claims 3, 8, 10, 14, 25 and 26 of the patent in suit.
These claims are as follows:
"3. A ferromagnetic iron oxide material adapted to
form an element of a magnetic impulse record member,
said material consisting essentially of acicular
crystalline particles uniformly small in size and not
over 6 microns in their greatest dimension of a
synthetic magnetic oxide of iron selected from the
group consisting of magnetic ferrosoferric oxide
Fe3O4, and magnetic gamma ferric oxide,
Fe2O3, the selected synthetic magnetic oxide
of iron having a cubic lattice structure, and said
material having a coercive force value of between 200
and 550 oersteds and a ratio of Bfm/Br at
H-1000 of not over 3 to 1.
"8. The method of making permanent magnet material
which comprises precipitating a non-magnetic ferric
oxide from solution in acicular crystalline form,
heating said non-magnetic ferric oxide in a reducing
hydrogen atmosphere to a temperature of about 750° F.
for a sufficient length of time to reduce said ferric
oxide to a magnetic ferrosoferric oxide, cooling the
ferrosoferric oxide in the presence of a reducing
atmosphere to room temperature and then exposing said
ferrosoferric oxide to the air to recover a
ferrosoferric oxide having a coercive force value
between 200 and 550.
"14. The method of making magnetic iron oxide which
comprises providing a synthetic non-magnetic ferric
oxide in acicular crystalline form and of particle
size not over 6 microns in its greatest dimension,
reducing said non-magnetic oxide at elevated
temperatures to produce a ferrosoferric oxide and
oxidizing said ferrosoferric oxide to gamma ferric
oxide having permanent magnet properties including a
coercive force of at least 200 oersteds.
"25. Ferromagnetic iron oxide selected from the
group consisting of a synthetic ferrosoferric oxide,
Fe3O4, and of a synthetic gamma ferric oxide,
Fe2O3, adapted to form an element of a
magnetic impulse record member, said iron oxide
consisting essentially of uniformly small elongated
crystals of less than about 1.5 microns maximum
dimension having a length-to-width ratio of about 2.5
to 1 and higher, and having a cubic crystal lattice
structure and a coercive force, Hc, within the
range of 245 to 330 and remanence, Br, of above
about 500 gauss.
"26. A magnetic impulse record member having a
non-magnetic carrier and a coating adherently bonded
thereto of a binder and magnetic material, said
magnetic material being the ferromagnetic iron oxide
defined in claim 25 and having a Br versus H
characteristic that rises most rapidly at fields
between 200 and 600 oersteds and relatively slowly at
fields between 0 and 200 oersteds and at fields above
The patent in suit relates to a permanent magnetic material and
to a method of making same. This permanent magnetic material is
then impregnated or coated on a non-magnetic carrier to produce a
magnetic impulse record member.
Magnetic recording was first discovered prior to the beginning
of the present century by Valdemar Poulsen, a Danish physicist.
He found that a mass of magnetizable material may be impressed
with impulses varying in intensity in successive adjacent
portions of the magnetizable mass which will retain these
impulses. It was learned that a sound wave could be converted
into a varying electric signal and it was further determined that
by energizing an electromagnet with the electric signal a
corresponding magnetic signal would be impressed on the record
member as the record member was caused to pass over the
electromagnet. It was further observed that when the same record
member was again moved past the same or a similar electromagnet,
an electric signal would be introduced into the electromagnet
corresponding to the magnetic signal which had been impressed on
the record member by its prior exposure to the effects of the
electromagnet. These early experiments were in great part
conducted using wire or metal tapes as the record member.
During the period referred to above the magnetic recording art
in the United States was limited to the use of wire or tape. Iron
oxides were not generally used as recording media in the United
States until after the conclusion of World War II when the Brush
Development Company of Cleveland, Ohio, placed on the market the
Brush "Soundmirror". This machine used as a recording media a low
coercive force magnetic oxide tape similar to that used by the
Germans in their Magnetophon and Tonschreiber. The basic
difference between the German and Brush machines was that the
Soundmirror operated at a speed of 7½ inches per second. This
speed has become the standard speed of American tape recorders.
While the work on the Soundmirror was progressing — it became
the first American machine for home use — the Indiana Steel
Products Company had produced a tape recorder designed to use
high coercive force tapes in order to obtain high frequency
response at low speeds. The tape used with this machine had a
coercive force of around 250 oersteds and utilized a magnetic
powder having particle sizes in the 1 micron range for the
recording media. The coating material used on this tape was not
the magnetic iron oxides claimed as an invention in the patent in
suit nor was it similar to that used on the German tapes.
On June 1, 1946 Marvin Camras, the inventor of the patent in
suit, applied for a patent on an "apparatus for magnetic
recording" and on October 31, 1950 Patent No. 2,528,261 issued
(Dx. N). In that patent he claimed as his invention a magnetic
recorder designed to use a recording media having a permeability
of less than 50 and a coercive force of over 100 oersteds,
preferably over 250 oersteds. With the advent of this magnetic
recorder the first great need for high coercive force and low
permeability tapes arose.
On July 25, 1947 Marvin Camras filed his application and on
November 16, 1954 the patent in suit was issued.
The important criteria of his patent, according to Mr. Camras,
(1) A relatively high coercive force, generally between 200 and
(2) Initial magnetization curve with a relatively gentle slope
to an Ho point of about 250 gauss.
(3) A rapid rise from about 200 to 600 oersteds.
(4) A high remanence above about 500 gauss and a Bfm/Br
ratio of less than 3 to 1.
(5) A synthetic material comprised of particles of preferably
less than 1.5 microns and not more than 6 microns in maximum
dimensions and consisting of acicular or elongated particles.
The defendants assert the invalidity of the patent and declare
that the patent should not have been issued.
One of the asserted grounds of invalidity is that the oxides
claimed as an invention by Camras are old in the art, as was also
the use of magnetic iron oxides for magnetic impulse record
"A suitable process for the preparation of the
oxides in a very finely divided state consists in the
decomposition of complex metal salts, as for example
iron ammonium salts, or complex metal compounds which
contain organic radicals such as pyridine. Such
methods of preparing oxides are known per se, but it
was hitherto not known that magnetic metal oxides
when prepared in such a way are highly suitable for
sound recording purposes * * *
"It has been found that the oxides to be used
according to this invention fulfil the said
requirements to a high degree. The main reason for
this is the extreme fineness and the very great
uniformity in the size of the particles. For example
it is possible to incorporate the magnetic material
according to this invention in the sound record
carrier in a particle size of 1/1000 millimetre or
less; with the processes hitherto proposed for the
purpose this is either impossible or only possible
with difficulty. The magnetic properties of the
oxides, in particular their induction and remanence
with low field strengths, are favourable for their
use as sound record carriers. The high coercive force
ensures a good stability of the sound recording. * *
"The following Examples will further illustrate how
the said invention may be carried out in practice but
the invention is not restricted to these Examples."
Johnson sets out five examples of methods of preparing oxides
for use in the invention. However, the patent clearly indicates
that the method of preparation of the oxide is not restricted to
The Patent Examiner rejected several claims of the Camras
patent as being unpatentable over the Johnson British patent (p.
48, Dx. G). To overcome this objection (p. 48, Dx. G) Camras
conducted experiments of three examples of the British patent. By
following these methods low coercive force materials were
obtained (pp. 63, 64, Dx. G).
On page 64 of this exhibit the effect of temperature upon the
magnetic properties of hydroxides was illustrated. Crystals of
Fe2O3.H2O in the form of light yellow acicular
crystals of a particle size from ¼ to about 1½ microns in length
and 1/10 to 3/10 microns in width were produced according to the
reactions described in the Camras application. The resulting
particles were then reduced in an atmosphere of hydrogen at
temperatures of 300° F., 400° F. and 1200° for varying periods of
time. The resulting oxides proved in the case of the 300° F. and
400° F. reductions to have coercive force values and remanence
values too low to obtain a measurement, and in the case of the
1200° F. reduction a coercive force of 30 was attained.
Since Camras in his experiments failed to follow the teachings
of the prior publications for the method of reducing both
synthetic and natural iron oxides, it is not surprising that he
obtained the low coercive force above indicated.
At the Inter-Partes demonstration at Easton, Pennsylvania, the
Williams and Thewlis publication was followed in conducting the H
series demonstration. In this demonstration the hydrate was
reduced at 350° C. (R. 2049). The particles of the oxide produced
in the H demonstration were found to be acicular and smaller than
1 micron in maximum dimension (R. 2051) (Dx. 1-B-1, p. 33). The
oxides produced in this demonstration showed a coercive force of
over 200 oersteds and a Bfm/Br ratio of less than 3 to 1
(R. 3166, 3167).
By following the teachings of the Williams and Thewlis
publication, as was done at the Inter-Partes H demonstrations, an
oxide clearly conforming to the patented oxide was obtained.
Another prior publication showing the effect of heat treatment
on magnetic properties of iron oxides is found in the Kraeber and
Luyken publication (Dx. 1-Z-1 and 1-Z-2).
This article was published in Germany in 1936 by the
German-Kaiser-Wilhelm Institute. It discloses methods by which
iron oxides may be reduced and oxidized in order to provide
certain magnetic properties. At the Easton Inter-Partes
demonstrations an oxide, identical to Camras' oxide, was produced
by following the disclosures of the publication.
The plaintiffs contend that the article fails to anticipate the
Camras patent, notwithstanding the fact that oxides identical to
those disclosed in the Camras patent were produced by following
the teachings of Kraeber and Luyken at the Easton demonstrations.
They assert that the publication does not disclose the use of
these oxides for magnetic recording media; that the Camras
measurements, which are necessary in order to determine that the
oxide sought has been obtained, are not disclosed anywhere in the
article; and that the starting material used at the Easton
demonstration was not the Merck iron hydroxide but rather was a
product of the defendant Williams Company. This latter objection
is strongly urged by the plaintiffs in their claim that the
Camras oxide was not anticipated by this publication.
The defendants used a substitute hydroxide (a commercial
product of defendant Williams known as YLO-1788); this iron
hydroxide corresponds to the Merck iron hydroxide prescribed as a
starting material in the publication. The Court finds that the
defendants have justified their use of the substitute hydroxide
by showing that:
(1) They attempted to obtain the Merck iron hydroxide as
prescribed by Kraeber and Luyken, but none was available. (p. 73,
(2) The substitute starting material (YLO-1788) was of the same
extremely small particle size as Merck iron hydroxide. The
evidence discloses that the particle size of the prescribed and
the substitute hydroxides are the same. (p. 11, Dx. 3-E) (pp. 16,
17, Dx. 1-Z (2)) (p. 14, Dx. 3-E).
(3) The Merck and the YLO-1788 hydroxides both consisted of
acicular particles. This is borne out by the evidence. (pp. 62
and 69, Dx. 3-E, and pp. 8, 16 and 17, Dx. 1-Z-2).
Except for the use of a substitute starting material, the
procedures outlined in the Kraeber and Luyken article were
followed in producing the P series demonstration oxides. The
starting material (p. 11, Dx. 3-E) was roasted at a temperature
of 400° C. for one-half hour in the presence of carbon monoxide
gas (pp. 8, 9, Dx. 3-E). The kiln was then removed from the
heating blanket with nitrogen continually passing through and was
placed on the floor to cool down to room temperature (pp. 18, 19,
Dx. 3-E). This
procedure follows test No. 3, table 5, page 30 of the Kraeber and
Luyken article (Dx. 1-Z (2), p. 8, Dx. 3-E). In the second step
in the procedure materials produced in the first step were placed
in a rotary kiln and heated to 200° C. This temperature was
maintained for a period of one hour as air passed through the
kiln. The kiln was then removed from the heating blanket and
cooled down to room temperature (p. 27, Dx. 3-E). The second step
of the procedure follows the procedure outlined in experiment 7,
table 6, page 36 of the publication (Dx. 1-Z(2)).
The oxide and tapes produced correspond in all respects to the
Camras patent material (pp. 2-5, Dx. 1-B-3) pp. 1-3, Dx. 1-B-4).
The United States Department of Interior Bureau of Mines
Bulletin No. 425, dealing with the Magnetic Separation of Ores,
at page 136, teaches the Effect of Heat Treatment on Iron Oxides,
Ferrites and Ilmenite and provides as follows:
"Properties of Gamma Fe2O3 and Ferrites
"It has long been known that under certain
conditions Fe2O3 is ferromagnetic. This form
has been shown by X-ray spectrometry to be cubic,
whereas the ordinary para-magnetic form is hexagonal.
"Of the various methods for preparing this
ferromagnetic Fe2O3 two may be used to treat
minerals for magnetic separation. By the first method
Fe3O4, regardless of its method of formation,
may be the starting point. Reactive varieties
resulting from the gaseous reduction of ordinary
Fe2O3 or from drying precipitated, hydrated
Fe3O4 are readily converted by heating in air
at 220° to 550° C. or by subjecting them to the
action of oxidizing solutions to gamma Fe2O3.
Heating at higher temperatures forms ordinary
Fe2O3. Nonreactive Fe3O4, exemplified
by most natural magnetities, may be converted to the
reactive type by oxidation to the ordinary
Fe2O3 followed by gaseous reduction.
"By the second method gamma ferric oxide hydrate is
heated at 250° to 300°C. (347) so as to produce the
ferromagnetic gamma Fe2O3. The transformation
temperature is not sharp, but overheating must be
avoided to prevent further conversion to ordinary
"Gamma ferric oxide hydrate may be formed by the
oxidation of ferrous hydroxide or of mixtures of
ferrous and ferric hydroxides (210); its preparation
may be favored by the use of certain organic
nitrogen-containing compounds (15)."
The teachings of the above article disclose to one skilled in
the art that in order to obtain a high coercive force oxide,
reduction temperatures of between 220 to 500°C. must be followed
and that overheating will destroy these magnetic properties.
In the demonstration Camras (Camras' affidavit, pp. 62, 63, 64,
Dx. G) conducted for the benefit of the Patent Examiner it is
important to note that the temperatures prescribed above were not
followed. Although the Bureau of Mines Bulletin No. 425 was used
as a reference by the Examiner in determining the validity of the
Camras application, only pages 88 to 95 were considered,
notwithstanding the fact that at page 136 the effect of heat
treatment on iron oxides is disclosed. Could it be that Camras
"religiously avoided" conducting an experiment based on this page
of the publication?
The Court finds that the Williams and Thewlis, the Kraeber and
Luyken, and the Bureau of Mines publications anticipated the
oxide claimed by Camras as his invention. The Camras file wrapper
(Dx. G) discloses that these publications were not cited to nor
considered by the Patent Office as prior art publications. The
Patent Examiner did not have the benefit of these publications
when he considered the validity of the Camras application.
The Court further finds that the Johnson British patent clearly
teaches the use of magnetic iron oxides, both Fe3O4 and
gamma Fe2O3, in a finely divided state and of a small
particle size and of a high coercive force to be especially
suited for use as a magnetic recording media. Although the
examples of the method of preparing the oxide contained in the
Johnson patent do not disclose the exact method of preparing the
oxide claimed by Camras as his invention, yet by following the
Johnson teachings and referring, as the patent teaches, to the
magnetic oxide art as known and disclosed in prior publications,
the Camras oxide is anticipated, as is the use of these oxides
for magnetic recording members.
The Welo and Baudisch article published in August, 1934, (Dx.
B) discloses a method whereby synthetic magnetite may be obtained
by the reduction of gamma monohydrate and alpha monohydrate.
The starting material for the gamma ferric oxide of Welo and
Baudisch is obtained by burning iron carbonyl in a plentiful
supply of air, then reducing the material in sodium acetate or
potassium nitrate at 320°C. for 10 minutes (p. 73, Dx. B). This
article was followed at the Easton demonstrations. Reduction at
460°C. in acetate resulted in the production of oxides having
Hc of over 200 oersteds, Bfm/Br of 3 to 1, rapid rise
from 200 to 600. Reduction at 360°C. in potassium nitrate
produced oxides having Bfm/Br of 3 to 1 and a coercive
force of 200. (Dx. 1-B-1, pp. 34, 35, 36).
Electron micrographs taken of the demonstration oxides produced
by following Welo and Baudisch show the presence of some
particles with a length to width ratio of more than 2 to 1 (Dx.
1-R-5 and 1-R-6).
When viewing the particles under the definition of acicular,
meaning 2 to 1 or more in length to width ratio, defendants'
Exhibits 1-R-5 and 1-R-6 show the presence of acicular particles.
However, the oxide produced by following Welo and Baudisch does
not produce predominantly acicular particles as does the Williams
IRN-210 oxides, claimed to infringe the patent.
Camras claims and places much stress on acicularity as one of
the important criteria of his oxide. It should therefore be
particularly noted and emphasized that he did not produce or
offer in evidence a photomicrograph which would show to what
extent the Camras oxide is more acicular than the oxides produced
at the Easton demonstrations, if in fact such is true.
Photomicrographs of the Welo and Baudisch oxides produced at the
Easton demonstrations are in evidence (Dx. 1-R-5, 1-R-6).
Photomicrographs of IRN-210 are also in evidence, having been
offered by the plaintiff (Px. 26).
An article by J. Huggett of the University of British Columbia,
published in 1929, discloses a method of precipitating
Fe2O3 by reacting a solution of ferric nitrate with
ammonia and then reducing the Fe2O3 in a mixture of
hydrogen and water vapor. The reference teaches how to control
the concentration of hydrogen and water vapor in relation to the
temperature so as to obtain ferrosoferric oxide ...