The Los Alamos Primer

The following notes are based on a set of five lectures given by

R. Serber during the first two weeks of April 1943, as an
“indoctrination course” in connection with the starting of the
Los Alamos Project. The notes were written up by E. U. Condon.

Everybody assembled in the big library reading room on the first

floor of the Technical Area, the building where the theoretical phys-
icists had their offices. We had a little blackboard set up in front and
a lot of folding chairs spread around the room. Fifty people on hand,
something like that. Scientific staff, a few visiting VIPs. There was
hammering off in the background, carpenters and electricians
working out of sight but all over the place. At one point during the
lectures a leg came bursting through the beaverboard ceiling. One
of the workmen misstepped and they had to pull him out.

1. Object

The object of the project is to produce a practical military

weapon in the form of a bomb in which the energy is released

[ 1 ]
by a fast neutron chain reaction in one or more of the materials
known to show nuclear fission.
I started lecturing. I started talking about the “bomb.” After a
couple of minutes Oppie1 sent John Manley up to tell me not to use
that word. Too many workmen around, Manley said. They were
worried about security. I should use “gadget” instead. In the Primer
Condon wrote it down both ways. But around Los Alamos after that
we called the bomb we were building the “gadget.”
Section 1 emphasizes that our purpose at Los Alamos was to
build a practical military weapon—one small enough and light
enough that an airplane could carry it.2 There was no use making
something that weighed one hundred tons. That was our concern.
We meant to build this weapon by utilizing the energy from
nuclear fission. Fission had a history. For a long time before 1939,
people were bombarding uranium with neutrons. Uranium was the
heaviest element known up to 1939. People had the idea that the ura-
nium they were bombarding was capturing neutrons and transmut-
ing to heavier elements, elements beyond uranium on the periodic
table, transuranics. I remember seminars in Berkeley in the 1930s
when the chemists discussed the trouble they were having explain-
ing the chemistry of these supposed transuranic elements. The
chemistry didn’t seem to be working out right. Then Otto Hahn and
Fritz Strassmann, in Germany, working with the physicist Lise
Meitner, found out that making transuranics wasn’t what was

1. J. Robert Oppenheimer.
2. On Edward Teller’s blackboard at Los Alamos I once saw a list of
weapons—ideas for weapons—with their abilities and properties displayed. For
the last one on the list, the largest, the method of delivery was listed as “Back-
yard.” Since that particular design would probably kill everyone on Earth, there
was no use carting it elsewhere.

[ 2 ] T h e L o s A l a m o s P r i m e r
usually happening at all. Instead, the uranium nucleus was actually
splitting into two big pieces, and doing it with the release of a great
deal of energy (and a couple of extra neutrons, as several people soon
demonstrated). As soon as that was discovered, everybody realized
the possibility both of making weapons and of getting power.
Finally, the reaction we were interested in was a fast neutron
chain reaction, which I’ll discuss later in these notes.

2. Energy of Fission Process

The direct energy release in the fission process is of the order of

170 Mev per atom. This is considerably more than 107 times the
heat of reaction per atom in ordinary combustion processes.
In Section 23 we immediately come to the heart of the matter:
that the energy released in the fission of the uranium nucleus is
considerably greater than 107—that is, ten million times the energy

3. In Section 2 we begin using so-called scientific notation. Ordinary deci-

mal notation is inconvenient when you’re dealing with very large or very small
numbers:

1,000,000 × 10,000,000 = 10,000,000,000,000

Ten followed by twelve zeros, ten trillion, isn’t an easy number to read. It’s sim-
pler and more convenient to tell how many zeros there are after the 1 instead
of writing them all down. Thus we write 10n to mean a 1 with n zeros after it.
10 is 101, 100 is 102, 1,000 is 103, and so on. Convert equation (1) to scientific
notation (substituting a dot for the multiplication sign) and it looks like this:

106 · 107 = 1013

with the nice advantage that we can do multiplication by simply adding the
superscripts, which in fact are powers of ten. We write 2,500,000 like this:

2.5 · 106

E n e r g y o f F i s s i o n P r o c e s s   [ 3 ]
released in a typical chemical combustion such as an explosion or
a fire. All else follows from this fact. So we should try to understand
where this large number comes from.
We can do so because the origin of the energy released in fis-
sion is exactly the same as the origin of the energy released when
two atoms or molecules react chemically. It’s the electrostatic
energy between two similarly charged particles. Two similarly
charged particles repel each other. There’s an electrical force push-
ing them apart. Work has to be done to overcome that repulsion
and push them together from a large distance, up to a point of sepa-
ration we can call R.
To start with a simpler particle than an atom, let’s look at two
electrons pushed together. If you released them, they would fly
apart with an amount of energy equal to the work that went into
pushing them together. That energy E is given by the formula

e2
E = R (1)

where e is the electron charge, e2 is e multiplied by itself, and R is the

distance between the particles. The electrostatic energy thus ends

This notation can be extended in turn to very small numbers by using a nega-
tive superscript, 10−n which means 1 divided by 10 to the n. Thus, 10−1 is 1/10th,
10−2 is 1/100th, and so on. In decimal notation, 10−1 = 0.1, 10−2 = 0.01, 10−3 =
0.001. To get

2.5 · 10n

you write 2.5 and move the decimal point n places to the right; to get

2.5 · 10−n

you write 2.5 and move the decimal point n places to the left.

[ 4 ] T h e L o s A l a m o s P r i m e r
up as kinetic energy, the energy of motion. In chemical reactions—
the burning of hydrogen and oxygen in a rocket engine, for
example—electrons bound in atoms or molecules change their
positions, and the change in electrostatic energy is what appears as
the energy of the chemical reaction.
Now let’s consider the electrostatic energy in the uranium
nucleus. The uranium nucleus contains 92 protons, each of which
has the same charge as an electron, though of opposite sign—
particles of opposite sign attract each other, those of the same
sign repel. So the uranium nucleus has a charge 92 times as great as
an electron; it’s positive rather than negative, + rather than −, but
since only the square of the charge is involved, that difference
doesn’t matter in equation (1). The numerator of (1) is thus
922 times bigger than for a chemical reaction. For our purposes,
922 is close enough to call 1002. So the numerator for a uranium
atom would be greater by a factor of 1002, 100 times 100, or
10,000 (104).
The uranium nucleus is also much smaller than an atom. In an
atom, the distance R is 10−8 cm (cm meaning centimeters). The
radius of the uranium nucleus is 10−12 cm, which is 104 times
smaller. The electrostatic energy for a uranium nucleus is therefore
104 for the numerator and another 104 for the denominator, for a
total of 108 times greater than the electrostatic energy between
atoms or molecules. When a uranium nucleus fissions, much of this
energy is released as kinetic energy in the two fission fragments
that fly apart. Suppose that the uranium nucleus broke in half. Each
fragment would have half the charge. The numerator of equation
(1) would be a quarter as big—a half times a half. Since the volume
is proportional to the cube of the radius, the radius would be
smaller by a factor of

E n e r g y o f F i s s i o n P r o c e s s   [ 5 ]
 3
1/ 2 = 1/1.26

So each fragment would have an electrostatic energy of about a

third of the total and the two fragments about two-thirds. That
leaves a third left over for the reaction energy.
Thus we see that the energy of fission is about 108—one
hundred million times—greater than the energy of a chemical
reaction, confirming the statement that it’s “considerably more
than 107.”
This is 170 · 106 · 4.8 · 10−10/300 = 2.7 · 10−4 erg/nucleus. Since
the weight of 1 nucleus of 25 is 3.88 · 10−22 gram/nucleus the
energy release is

7 · 1017 erg/gram

The energy release in TNT is 4 · 1010 erg/gram or 3.6 · 1016 erg

/ton. Hence

1 kg of 25 ≈ 20000 tons of TNT

To compare the energy released per gram of uranium versus a gram

of a chemical explosive such as TNT, we have to remember that an
atom of uranium weighs ten times as much as the atoms involved
in the chemical reaction. So in a given weight of uranium there will
only be a tenth as many atoms. We have to reduce our figure of 108
to 107 to compare equal weights of uranium and chemical explo-
sive. That means that one kilogram of uranium, if it fissioned com-
pletely, would be equivalent to about 104 tons of explosives—10,000
tons, 10 kilotons, which is reasonably close to the actual figure at

[ 6 ] T h e L o s A l a m o s P r i m e r
the end of Section 2 of 20,000 tons. (Twenty thousand may not
look “reasonably close” to 10,000 if you’re not used to thinking in
terms of “orders of magnitude,” which are factors of 10. Ten thou-
sand and 20,000 are of the same order of magnitude, 104; one is
1 · 104 and the other is 2 · 104.)
Somehow the popular notion took hold long ago that Einstein’s
theory of relativity, in particular his famous equation E = mc2, plays
some essential role in the theory of fission. Albert Einstein had
a part in alerting the United States government to the possibility
of building an atomic bomb, but his theory of relativity is not
required in discussing fission. The theory of fission is what physi-
cists call a nonrelativistic theory, meaning that relativistic effects
are too small to affect the dynamics of the fission process
significantly.
Section 2 of the Primer gives a more exact calculation of the
ratio of the energy released by the fission of a gram of uranium to
the energy released by the explosion of a gram of TNT. To get the
ratio of such quantities, you have to measure them in the same
units. That complicates things, because in different branches
of science it’s convenient to use different units to measure the
same quantity. A chemist is likely to measure energy in calories,
while the standard unit of energy for the physicist is the erg.
The erg is rather too small a unit to be convenient for everyday
use. Utilities bill customers for kilowatt hours of electric energy;
there are 3.6 · 1013 ergs in a kilowatt hour. On the other hand,
the erg is too large a unit to be convenient for an atomic physicist,
who uses a smaller and different unit, the electron volt: the
energy acquired by an electron falling through a potential
difference of one volt. That’s a convenient size; the energy that

E n e r g y o f F i s s i o n P r o c e s s   [ 7 ]
binds an electron in a hydrogen atom, for example, is just 14 ev—14
electron volts. The energy of typical chemical bonds is just a few ev.
The nuclear physicist has borrowed the unit and uses it in larger
multiples: Kev, meaning 1,000 ev (103); Mev, meaning 1,000,000
ev (106).
The Primer gives the energy released in fission as 170 Mev. To
compare this number with the energy released by TNT, which is
given in ergs per gram, you have to know how many electron volts
there are in an erg. The simplest and most reliable way to answer
this question is to go to a library and take down a reference book
like the Handbook of Chemistry and Physics, which has elaborate
tables giving the ratios of various units of measurement (thus 12
inches equals 1 foot, an inch equals 2.54 centimeters, an ounce
equals 28 grams—these are the sort of ratios I mean). But at Los
Alamos at this time, in April 1943, although we had a librarian—
my wife Charlotte—and a library, we didn’t yet have library
books. So apparently I didn’t answer the question the easy way. I
figured out the ratio on the back of an envelope using the definition
of an electron volt and some numbers I remembered. This is the
mysterious little calculation that begins the second paragraph of
Section 2:

170 · 106 = energy in ev

times
4.8 · 10−10
which is the charge on the electron
/300
since a volt is 1/300th of the electrostatic unit of voltage
= 2.7 · 10−4 erg/nucleus

[ 8 ] T h e L o s A l a m o s P r i m e r
which is the equivalent in ergs to the energy in ev. My calculations
indicated that the energy of fission of 1 kg of uranium equalled
20,000 tons of exploding TNT—the wiggly equal sign (≈) means
“approximately equal to.” When I referred to tons I meant short
tons, by the way: 2,000 pounds, or 907 kilograms.
In Section 2 I refer to the rarer form of uranium, the form we
were interested in, as “25.” This simple code was commonly used
in the Manhattan Project; 25 meant U235, 28 meant U238, 49 meant
one kind of plutonium, Pu239. Uranium is element 92, plutonium
element 94, the numbers referring to the number of protons in
their nuclei. The sum of both protons and neutrons in the nuclei of
atoms gives what is loosely referred to as the atomic weight (it’s not
really a weight). Every kind of uranium has 92 protons, every kind
of plutonium has 94 protons, but different kinds differ in their
numbers of neutrons. These different kinds are called “isotopes.”
The common isotope of uranium has atomic number 92 and atomic
weight 238 (indicating 146 neutrons: 238 total nucleons — 92 pro-
tons = 146). A much rarer isotope, the one we were interested in,
has atomic number 92 but atomic weight 235. The isotope of pluto-
nium we were interested in has atomic number 94 and atomic
weight 239. The code we used simply took the last digit of the
atomic number and put it together with the last digit of the atomic
weight: 92238 became 28, 94239 became 49, and 92235, as here in
Section 2, became 25. Since our work on the atomic bomb was a
military secret, we weren’t supposed to say the words “uranium”
and “plutonium” aloud. That’s why we used the code.
Twenty thousand tons is a pretty impressive figure for one kilo-
gram of anything. Seven times 1017 ergs per gram is nearly 20,000
kilowatt hours. So one pound of uranium, 454 grams, would release

E n e r g y o f F i s s i o n P r o c e s s   [ 9 ]
9 million kilowatt hours, for which my local electric utility,
Consolidated Edison, would charge me more than one and a quar-
ter million dollars.

3. Fast Neutron Chain Reaction

Release of this energy in a large scale way is a possibility

because of the fact that in each fission process, which requires
a neutron to produce it, two neutrons are released. Consider a
very great mass of active material, so great that no neutrons
are lost through the surface and assume the material so pure
that no neutrons are lost in other ways than by fission. One
neutron released in the mass would become 2 after the first
fission, each of these would produce 2 after they each had
produced fission so in the nth generation of neutrons there
would be 2n neutrons available.

Having established roughly how much energy might be available

from fissioning a quantity of uranium, I next began discussing how
to get this energy out.
Massive energy release from fission depends on developing a
chain reaction—a geometric progression of fission events, one trig-
gering two, two triggering four, four triggering eight, and so on.
That phenomenon depends in turn on the propensity of what the
Primer calls “active materials”—U235 and Pu239, for example—to
eject more neutrons per fission on the average than they absorb
when they’re bombarded with neutrons. Enrico Fermi, Frederic
Joliot, Leo Szilard, and others found secondary neutrons from fis-
sion in experiments they conducted independently, within days of
each other early in 1939, in Paris and New York. This first paragraph

[ 10 ] T h e L o s A l a m o s P r i m e r
of Section 3 assumes an ideal arrangement of material where no
neutrons are lost through the surface or to impurities. Fission of
U235 releases 2.2 secondary neutrons on the average; 2 is a reason-
able order-of-magnitude rounding of that number.
Since in 1 kg. of 25 there are 5 · 1025 nuclei it would require
about n=80 generations (280 ≈ 5 · 1025) to fish the whole
kilogram.
The second paragraph of Section 3 is notable for a mistake.
There are not 5 · 1025 nuclei in a kilogram of uranium. There are
2.58 · 1024. Uranium metal has a density of 19 grams per cubic cen-
timeter; 5 · 1025 is 19 times 2.58 · 1024 and is thus the number of nuclei
in 1,000 cubic centimeters, not 1,000 grams. On the other hand, 280
is not 5 · 1025 but 1.2 · 1024. So 80 generations is still the right answer
(81 if you want to be cranky about it). Since fission occurs in about
10−8 seconds, those 80 generations would pass in .8 microseconds:
it would take less than a millionth of a second to fission a kilogram
of uranium.
In these notes I use the verb “to fission.” In the Primer we used
the verb “to fish.” That’s some indication of how new our work was.
Otto Frisch and Lise Meitner named the new nuclear reaction they
confirmed in 1939 “fission,” borrowing the word from biology. We
hadn’t settled on a verb form of the noun yet. “To fish” didn’t stick.
Today we say “to fission,” but we kept the pronunciation: it’s
“fishin’,” not “fizj-un.”
While this is going on the energy release is making the
material very hot, developing great pressure and hence tend-
ing to cause an explosion.
The statements in Section 2 tend to be laconic. If the reaction
proceeded at 10 percent efficiency, it would heat the uranium ini-
tially, in less than a millionth of a second, to a temperature of about

F a s t N e u t r o n C h a i n R e a c t i o n   [ 11 ]
1010 degrees Celsius—about 10 billion degrees. The pressure devel-
ops accordingly, and the explosion is correspondingly powerful.
In an actual finite setup, some neutrons are lost by diffu-
sion out through the surface. There will be therefore a certain
size of say a sphere for which the surface losses of neutrons are
just sufficient to stop the chain reaction. This radius depends
on the density. As the reaction proceeds the material tends to
expand, increasing the required minimum size faster than the
actual size increases.
The whole question of whether an effective explosion is
made depends on whether the reaction is stopped by this ten-
dency before an appreciable fraction of the active material has
fished.
As the sphere expands, the density of the material within it
drops, which simply means that the atoms are further apart. The
distance a neutron moves between nuclear collisions increases and
as a result more neutrons escape through the surface before mak-
ing another fission. As the expansion proceeds, more and more
neutrons escape, until the loss is enough to stop the chain reaction.
This process is described in more detail in Section 13.
Note that the energy released per fission is large compared
to the total binding energy of the electrons in any atom. In con-
sequence, even if but ½% of the available energy is released
the material is very highly ionized and the temperature is
raised to the order of 40 · 106 degrees. If 1% is released the
mean speed of the nuclear particles is of the order of 108 cm/
sec. Expansion of a few centimeters will stop the reaction, so
the whole reaction must occur in about 5·10−8 sec otherwise the
material will have blown out enough to stop it.

[ 12 ] T h e L o s A l a m o s P r i m e r
 Now the speed of a 1 Mev neutron is about 1.4 · 109 cm/sec
and the mean free path between fissions is about 13 cm so the
mean time between fissions is about 10−8 sec. Since only the
last few generations will release enough energy to produce
much expansion, it is just possible for the reaction to occur to
an interesting extent before it is stopped by the spreading of
the active material.
It should be realized that at temperatures of tens of millions of
degrees the uranium is no longer a metal but has been converted to
a gas, a gas at tremendous pressure which will expand very rapidly.
We can estimate the velocity of expansion for 1 percent energy
release from the relation

E = ½Mv2 (2)

where E is energy, M mass and v velocity. Using the figures for the
energy released per fission and for the mass of a uranium atom
given in Section 2, we do a little calculation:

E = 1% of fission energy
= .01 · 2.7 × 10−4 = 2.7 · 10−6 ergs
M = 3.88 · 10−22 gm
v2 = 2E/M = 1.4 · 1016
v = 1.2 · 108 cm/sec (3)

and we find that the velocity of the nuclear particles would indeed
be about 108 cm/sec. This estimate assumes that all the energy is
transformed into energy of expansion, which is not literally true but
is an adequate assumption for an order-of-magnitude estimate. In

F a s t N e u t r o n C h a i n R e a c t i o n   [ 13 ]
any event, the velocity of expansion can’t be greater than the
number we’ve derived. That I based my calculations on the assump-
tion of releasing only 1 percent of the fission energy indicates that
in 1943 we would have been satisfied with quite low efficiencies.
In these paragraphs we also run into the technical term “mean
free path.” Since the concepts of mean free path and cross section
are essential to the rest of the discussion, they need to be explained.
Both concern the likelihood that a neutron will encounter and fis-
sion a uranium atom. (For a more detailed technical discussion, see
endnote 1.)
The mean free path is a number derived by measurement: the
distance a neutron traveling through a mass of material such as
uranium moves, on the average, before colliding with a nucleus of
that material.
Cross section is the area of the nucleus, πR2 (3 · 10−24 cm2). This
is the area that the neutron has to hit, the geometrical cross section.
When a neutron strikes a uranium nucleus, it’s temporarily
absorbed to make a nucleus with one extra neutron. Then one of
several things can happen. A certain fraction of the time, the com-
bined nucleus fissions, with a corresponding release of energy and
ejection of secondary neutrons. A certain fraction of the time a
neutron is emitted with lower energy than the original neutron, a
process called inelastic scattering.
The fission cross section is the fraction that leads to fission times
the geometrical cross section (that is, times πR2). The inelastic cross
section is the fraction that leads to inelastic scattering times the
geometrical cross section. The sum of these numbers adds up to
the geometrical cross section.
But this description is not quite exact. I’ve been discussing a
purely geometrical picture of the nucleus. There is a quantum

[ 14 ] T h e L o s A l a m o s P r i m e r
mechanical effect which causes the path of a neutron that just
misses the edge of a nucleus to be bent. The neutron comes out
with unchanged energy but in a different direction. This is called
elastic scattering and the cross section for its occurrence is called
the elastic cross section. The total cross section, the sum of the fission,
inelastic and elastic cross sections, will thus be somewhat bigger
than the geometrical cross section.
Slow neutrons cannot play an essential role in an explosion
process since they require about a microsecond to be slowed
down in hydrogenic materials and the explosion is all over
before they are slowed down.
The last paragraph in Section 3, concerning slow neutrons, will
be clearer after we look at figure 1 in Section 4 of the Primer. Let’s
skip it for now and return to it then.

4. Fission Cross-sections
235 238
The materials in question are U92 =25, U92 =28 and element
94239 =49 and some others of lesser interest.
Ordinary uranium as it occurs in nature contains about
1/140 of 25, the rest being 28 except for a very small amount
of 24.
When I reread the first sentence of Section 4 I was struck by the
phrase “element 94239 =49” where the structure of the sentence
seemed to demand “Pu =49.” I checked and discovered that the
word “plutonium” is never used in the Primer. Glenn Seaborg pro-
posed the name in 1942. I wonder if I was aware of it yet in April 1943.
The second paragraph of Section 4 conceals a very great effort
of human enterprise. In order to make an atomic bomb with ura-
nium the United States had to separate the 1/140th part of

F i s s i o n C r o s s - s e c t i o n s   [ 15 ]
U235 from the 139 parts of U238 in natural uranium when the only
difference between the two for purposes of separating them was
their mass. Most of the two billion dollars that the wartime pro-
gram to develop the atomic bomb—the Manhattan Project—spent
was invested in building the vast machinery necessary to separate
uranium. One system, gaseous diffusion, converted natural ura-
nium to a gas and then relied on the two isotopes’ differing rates of
diffusion across a porous barrier to accomplish the separation, but
the difference is so slight it required a cascade of several thousand
barrier tanks, the largest of them 1,000 gallons in volume, to enrich
the product to bomb grade. The building that held the gaseous-dif-
fusion plant at Oak Ridge, Tennessee, was correspondingly large—
a U-shaped structure with each leg of the U nearly half a mile long.
Another system, electromagnetic separation, relied on the fact that
an electrically charged atom traveling through a magnetic field
moves in a circle at a radius determined by its mass. Ions of a vapor-
ous uranium compound projected through a strong magnetic field
inside a curved vacuum tank separate into two beams, with lighter
U235 atoms following a narrower arc than heavier U238 atoms. Metal
pockets set at the end of the thousands of tanks built at Oak Ridge
collected each beam of isotopes separately in the form of metal
flakes. The system was notoriously inefficient, but it got the job
done. Most of the uranium used in the Hiroshima bomb was sepa-
rated this way.
Another great effort was required to produce plutonium. This
element does not occur naturally but has to be manufactured in a
nuclear reactor. In the reactor, fission neutrons are slowed down in
graphite (carbon) and some of them are captured in U238 to produce
the isotope U239 (since a neutron is added, the atomic weight

[ 16 ] T h e L o s A l a m o s P r i m e r