Myth #1: The Maxwellian Velocity Distribution can be measured directly with a flux of a model gas.
In many physics laboratories in schools and universities, the experiment
"Maxwellian Velocity Distribution" can be found with a model gas of spheres and
measuring the velocity distribution at a flux of the model gas:
The documentation of that product can be found
here and examples can be found
REALITY: The ensemble of the model gas has (nearly) a Maxwellian Velocity Distribution
const1 *v2 *econst2 *v2 .
But only the flux of the pushed out balls is being measured in this experiment,
not the ensemble!
Because a physical flux is, by definition, the integral over the
density times velocity (v) normal to the cross section, the
velocity distribution of the pushed out balls is the Maxwellian Distribution
times v, also called Modified Maxwellian Distribution
const3 *v3 *econst2 *v2 .
This multiplication with v is also necessary for calculating the flow rate .
The real velocity distribution is more narrow than the Maxwellian
distribution and it's maximum is at a little higher speed; that's
the reason why the real velocity distribution is called modified
The correct name of the experiment would therefore be "Modified Maxwellian
This is also described and verified with many experimental results, e. g. in the
journal Physical Review, Volume 99, Number 4, August 15, 1955, pages 1314-1321,
article "Velocity Distributions in Pottasium and Thallium Atomic Beams" from
R. C. Miller and P. Kusch.
See also the book "Fundamentals of Statistical and Thermal Physics" from
Frederick Reif, Chapter 7.12 (Effusion).
In the german version of this book, "Statistische Physik und Theorie der Wärme", from
Frederick Reif, 3. Aufl., ISBN 311011383X, you can find it on page 322.
The reason of the misconcept that the Maxwellian Velocity Distribution can
be measured directly with a flux of a model gas has been caused in
the middle of the 20th century by some layman, e. g. with the
(wrong) article "Modellversuch zur kinetischen Theorie" in
Beiträge zum mathematisch-naturwissenschaftlichen Unterricht, Heft
8, März 1962, page 1-22, at page 14.
And others have copied the error without a check.
There are many examples for similar copy and paste errors, e. g. the
length of the Rhine, wich is 1233 km, but almost every source
used a length of 1320 km (+87 km) from the middle of the 20th century till
2010, because someone made an error which was copied without a check:
Another example is the legend of the "very high iron content" of
spinach with a value of 35 Milligramm / 100 g, which is only one tenth
Myth #2: The three-body problem is insoluble.
In many physics books and scripts we can find the statement that the
three-body problem is insoluble (except of only few special cases),
because Poincaré said so in the late 1800s.
Examples can be found
Poincaré did only show that with the mathematical methods he used,
an analytic solution for 3-body problem can not be found.
Several time later, in 1912, the necessary mathematical methods where developed
and the Finnish mathematician Karl Fritiof Sundman
proved that there exists an analytically solution:
Sundman's theorem for the 3-body problem.
This result was generalised to the case of n>3 bodies by Q. Wang in the 1990s:
The global solution of the n-body problem.
Myth #3: The second law of thermodynamics is necessary for defining the
arrow of time.
In many physics books and scripts we can find the statement that the
second law of thermodynamics and entropy are necessary even for the
Arrow of time.
Examples can be found
It's true that an arrow of time can be defined via the second law,
but only statistically. And since 1998 it is known that time-reversal symmetry is
broken in neutral kaons, which means that neutral kaons do show the arrow of time:
Since 2012 the
Time-Reversal Violation in the B0 Meson System
It is also known that several other subatomic interactions involving the weak nuclear
force do violate the conservation of parity. According to
Theorem, this means they are also time irreversible.
Long before 1998 it was known that Quantum mechanics defines an arrow of time
And there are other quantum mechanic effects wich define an arrow of time e. g. the by deliquescence
of a free particle because his (non-relativistic) wave function is:
|ψ(x, t)|2 = 1/(d *sqrt(2 *π(1+Δ2)))
Δ = t *h/(4*π*m*d2) .
This Gaussian wave packet
has the position average
and the position uncertainty
in non-relativistic Quantum mechanics.
So in Quantum mechanics the motion of the center of gravity of a free particle
is time-reversal but the deliquescing is not because it depends on
Therefore time-reversion, which means changing t to -t does not
change the deliquescing, because t2=(-t)2 .
Another point is that, because of the Uncertainty principle, the reversal of the deliquescing
is not possible in Quantum mechanics.
That's why the deliquescing of the wave function of a free particle defines a
Quantum mechanic arrow of time.
That's also a reason why it's hard to build a quantum computer.
Another example of a time irreversible process in Quantum mechanics is
Wave function collapse.
In classical physics there are similar time arrows. See for example the
Optical isolator (Rayleigh Lighttrap),
Myth #4: Sometimes a kilo is 1000 and sometimes it is 1024.
A kilo, abbreviated k, usually means 1000 but it is often used as the factor 1024, e. g.
here and for the file size under a Microsoft operating system.
Another example is this list of definitions, from
a) The kilo is and has always been a decimal prefix which is 1000 by definition:
http://en.wikipedia.org/wiki/Kilo and there has never been
a standard or legal law which says something else.
Since 1954 the kilo, and the other decimal prefixes (mega, giga, tera etc.) are part of the International System of Units (SI) and they can also be found as
decimal (not binary) prefixes in many laws, e. g. in the german
"Ausführungsverordnung zum Gesetz über die Einheiten im
Messwesen und die Zeitbestimmung", 2000.
Therefore using decimal prefixes as binary prefixes was always wrong and illegal.
Before the international standard SI, the kilo has been
used long before e. g. for the Kilogram and the kilo was
defined by the metric (not binary) system, more than 200 years ago in the proclamation
on June 22, 1799:
Since 1999 the international standard
IEC 60027-2 is the official and legal standard for the
and defines kibi, abbreviated ki, as 1024.
This standard also defines mebi (=1,048,576), gibi (=1,073,741,824), tebi
So the binary prefixes are well-defined by a worldwide standard since more than a decade!
Before IEC 60027-2 and the kibi, 1024 was written as capitalized prefix K,
not k, to make clear that it is NOT a kilo (because it was simply "K" and nothing
else); setting k to 1024 is therefore completely
An example with the capitalized prefix K is the output of the df command under
|Filesystem || 1K-blocks || Used
|| Use% || Mounted on
|/dev/hda4 || 332077476|| 201822772|| 130254704|| 61%|| /
|udev || 907756|| 152|| 907604|| 1%|| /dev
|/dev/hda2 || 29301628|| 6730332|| 22571296|| 23%|| /tmp
|/dev/hda3 || 19542436|| 3769236|| 15773200|| 20%|| /var
But that was not a good solution, because the "K" has never been a part of a standard or
legal law and the Kelly-Bootle Standard Unit (=1012) uses the same
Although for exact countable integer things like bits and bytes
it does not make sense to say 1024 is nearly 1000,
in many old software and even in 2009 actual
operating systems from Microsoft you can find the bug that binary prefixes
are confused with decimal prefixes but hardware manufacturers
(usually) do use the prefixes
correct. An example is the capacity of this old 120 GB IDE hard disk drive, where you can
find the correct equation
1 GB=1,000,000,000 Bytes
in the middle of the sticker:
Another example is the HDD Installation Guide from Samsung in 2009:
It's simple to use binary prefixes when you mean binary prefixes and
to use decimal prefixes when you mean decimal prefixes, but because
some people can't do it and create such crazy things like the 1.44 MB
where 1000*1024 was used for
mega and not 10002 (or 10242).
A remaining problem is that the IEC 60027-2 only defines binary prefixes greater
one, because decimal prefixes smaller one are often used and even the good old
ANSI-C (C99) has hexadecimal floating-point numbers.
An example is the output of an Analog-to-Digital-Converter which is the
multiple of 1/1024 Volt: It`s easy to say the output is n Mibivolt (not
Millivolt) but it`s not so easy to say the output is n*0.9765625 Millivolt.
So it makes sense to use
mibi, abbreviated mi, for 2-10
mikbi, µi, for 2-20
nabi, ni, for 2-30
pibi, pi, for 2-40
fembi, fi, for 2-50
atbi, ai, for 2-60
zepbi, zi, for 2-70
yokbi, yi, for 2-80
This makes more sense because the end of the name is always bi for binary.
So the complete table would be:
Another problem is that the Microsoft operating systems have
additional errors, beside the error KB = 1024 B and MB = 1024 KB,
as can be seen in the next picture from 2009:
As can be seen in this picture, under MS-Windows XP
the newer linux.txt is reported with 238 KB on the disc, with 237 KB in the
dialog, and the older linux.txt is reported with 233 KB on the disk and 232 KB
in the dialog.
So the (reported) file size under MS-Windows XP is context-dependent: In a
dialog the files are about 0.5 % smaller, even if they are shown with the
same unit (KB which means kiB under MS-Windows XP); the unit "K" is not constant
under Microsoft Windows XP.
The linux kernel is much better and reports
[151765.683387] sd 14:0:0:0: [sdc] 62333952 512-byte logical blocks: (31.9 GB/29.7 GiB)
for a MircoSD card (2015-11-10), with binary and metrical prefixes.
Myth #5: A minute always has 60 seconds
The minute is often defined as 60 seconds. Examples can be found
Because of leap seconds, a minute has 59, 60, 61 or 62 seconds:
An example was the leap second 2009-01-01 00:59:60.
That's the reason why the minute and greater time units are not constant and
therefore no SI unit.
Because of leap seconds 23:59:60, 23:59:61, 00:59:60 and 00:59:61 are generally valid times,
although two (positive) leap seconds in a row are defined but have never been used.
Other Science Myths pages:
"Science Myths" in K-6 Textbooks and Popular culture
The Misappliance Of Science
TEN MYTHS OF SCIENCE: REEXAMINING WHAT WE THINK WE KNOW...
Physics Myths and Physics Facts
Urban legends at Wikipedia
Mythbusters (Wikipedia page)
WEIRD SCIENCE at Scientific American (Fact or Fiction and Strange but True)
Sollen wir täglich 2,5 Liter Wasser trinken? Auch Ärzte glauben manchen Unsinn.
Ernährungsmythen - entzaubert
Fun Science Myths pages (fun fiction):
Other Science pages about facts and myths
Usenet Physics FAQ
Some (german) books about myths and facts
Wer nichts weiß, muss alles glauben
Lexikon der populären Irrtümer
Lexikon der populären Ernährungsirrtümer
Lexikon der Öko-Irrtümer
Das Lexikon der Großstadtmythen
Kurzes Handbuch der Quacksalberei
Die Ratte in der Pizza und andere moderne Sagen und Großstadtmythen
Der Elefant auf dem VW und andere moderne Sagen und Großstadtmythen
Lexikon der Rechtsirrtümer
Neues Lexikon der Rechtsirrtümer
Das dritte Lexikon der Rechtsirrtümer
Mythen des Alltags
Buch-Serie Moderne Legenden im Test
Ulrich und Johannes Frey: Fallstricke, Die häufigsten Denkfehler in Alltag und Wissenschaft
Verlag C.H. Beck, München 2009
240 Seiten, 12,95 Euro