Comment. acad. sc. Petrop. 7. 1740 (p.

124 onwards) : 1
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

Concerning the sums of series of reciprocals.

Leonhard Euler.

§. 1.
Now the series of the reciprocals of the powers of the natural numbers have been
explored and investigated to such an extent that it hardly seems likely that it would be
possible to find anything new about them. Indeed whichever of the sums of the series
have been considered, and nearly all those have involved an investigation of this kind of
series, still no suitable way has been found for expressing them. Now I also, as more
often than not, have treated a number of methods of summation, and I have diligently
pursued these series, yet I have never understood any other method by which I could
define the sum of these, other than as an approximation, or perhaps I could reduce the
sums to the quadrature of curves of quickest descent; I have set out these methods here in
the preceding read dissertations [i.e. in the Commentaries of the weekly seminars at the
St. Petersburg institute]. But here I discuss series of fractions, the numerators of which
are 1, and the denominators are either squares, cubes, or any other powers of the natural
numbers : series of this kind are 1  14  19  16
1  1  etc. , likewise
25
1  18  27
1  1  etc. and similar higher powers, the general term of which is expressed in
64
the form 1n .
x
§.2. Moreover recently in an unexpected manner, I have been able to deduce an
expression for the entire sum of this series 1  14  19  16
1  1  etc. , which depends on
25
the quadrature of the circle, thus, as if truly the sum of this series is obtained, then
likewise the quadrature of the circle follows. For I have found that the sum of this series
is a sixth part of the square of the periphery of the circle of which the diameter is 1; or
with the sum of this series put equal to s, it has the ratio 6 s to1 of the periphery to the
diameter. Moreover I will soon establish that the sum of this series is approximately 1.
6449340668482264364, and from six times this number, if the square root is extracted,
the number 3.141592653589793238 is indeed produced, expressing the circumference of
the circle the diameter of which is 1. Again, by following the same steps again by which I
have pursued this sum, the sum of this series beginning1  16 1  1  1  1  etc. , can
81 256 625
also be grasped from the quadrature of the circle. Now the sum of this series multiplied
by 90 gives the fourth power of the circumference of the circle of which the diameter is
1. And by similar reasoning also the sums can be determined of the following series, in
which the exponents of the powers are even numbers.
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 2
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

§.3. In which therefore, I explain most conveniently in an orderly manner the whole
argument of how I have arrived at this sum, making use of the circle. In the circle
AMBNA described with centre C and radius AC
or BC = 1, I consider any arc AM, the sine of
which is MP and the cosine truly CP. Now with
the arc AM put equal to s, with the sine PM = y,
and with the cosine CP = x, by the well-enough
known method, so the sine y and the cosine x can
be defined for the given arc s by series, it is
indeed, as can be considered,
y  s  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc. and
3 5 7

x  1  1s.2  1.2s.3.4  1.2.3s.4.5.6  etc.

2 4 6

Clearly from a consideration of these equations,

the sums of the above series of reciprocals can be
reached ; indeed each of these equations is directed almost towards the same target, and
on this account it suffices only for the one to be treated in the manner that I now present.

§.4. Hence the first equation y  s  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc. expressed the
3 5 7

relation between the arc and the sine. Whereby from this equation, from the given arc the
sine of this arc can be determined, as from the given sine, the arc of this can be
determined. Moreover, I consider the sine y as given, and I investigate how the arc s
ought to be elicited from the given sine y. Now here before all else it is to be observed,
that innumerable arcs correspond to the same sine y, hence the proposed equation for the
sine must present innumerable arcs. If indeed s is considered as the unknown in that
equation then it has infinite dimensions, and thus it is not to be wondered at, if that
equation should contain innumerable simple factors, any of which put equal to zero must
give a suitable value for s.

§.5. Moreover since, if all the factors of this equation are to become known, then all the
roots of this also or the values of s are to be known, and thus in turn if all the values of s
can be assigned, then also all the factors of this can be found. Moreover, so that I can
indicate both the roots and the factors better in this equation, I change the proposed
equation into this form : 0  1  s  s3  s5  etc. Now if all the roots of this
y 1.2.3.y [Link].5.y
equation, or all the arcs which have the same sine y, are A, B, C, D, E, etc. then the
factors also are all these quantities, 1  As , 1  Bs , 1  Cs , 1  Ds , etc. On account of which,

0  1 s  s3  s5  etc.  ( 1  As ) ( 1  Bs ) ( 1  Cs ) ( 1  Ds ) etc.
y 1.2.3.y [Link].5.y
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 3
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

§.6. Moreover, from the nature and resolution of the equations it is agreed that the
coefficient of the terms in which s is present, or 1y , is equal to the sum of all the

coefficients of s in the factors or, 1y  1A  B1  C1  D

1  etc. Then in turn the coefficient

of s2 is equal to 0, on account of this term being missing in the equation, is equal to the
sum of the factors taken two terms at a time of the series 1A , B1 , C1 , D
1 , etc. Again,

 1.21.3. y is the sum of three [four in the original ms.] terms of the series 1A , B1 , C1 , D
1 , etc.

In a like manner the sum of the factors taken four terms at a time of the same series is
equal to zero, and  [Link].5. y is the sum of the factors taken five terms at a time of the
series, and henceforth.

§.7. But with the smallest arc AM = A put in place, the sine of this is PM = y, and the
half-circumference of the circle is put equal to p, then we have
A, p  A, 2 p  A, 3 p  A, 4 p  A, 5 p  A, 6 p  A etc., and likewise
 p  A,  2 p  A,  3 p  A,  4 p  A,  5 p  A, etc.,
[i. e. the places where the sine is zero, both in a clockwise and in an anticlockwise sense.]
taken together are all the arcs having the same sine y. Therefore since before we assumed
the series 1A , B1 , C1 , D
1 , etc. that is changed into this :

1, 1 , 1 , 1 , 1 , 1 , 1 , 1 , 1 etc.
A p  A  p  A 2 p  A  2 p  A 3 p  A 3 p  A 4 p  A  4 p  A
Hence the sum of all these terms is equal to 1y ; but the sum of the factors taken two at a
time of this series is equal to 0 ; the sum of the factors from the three terms is equal to
1 , the sum of the factors from four terms is equal to 0; the sum of the factors from
1.2.3. y
five terms is equal to 1.2.3.41.5. y ; the sum of the factors from six terms is equal to 0. And
thus henceforth.

§.8. But if any series is had a  b  c  d  e  f  etc. the sum of this is α, the sum of the
factors from the terms taken two at a time is equal to β; the sum of the factors from the
terms taken three at a time is equal to γ; the sum of the factors taken four at a time is δ,
etc. then the sum of the squares of the individual terms , that is
a 2  b 2  c 2  d 2  e 2  f 2  etc. =  2  2 ;
now the sum of the cubes a 3  b3  c3  d 3  e3  f 3  etc. =  3  3  3 ; the sum
of the fourth powers  4  4 2   4  2 2  4 .
Moreover so that it is appears clearer, as these formulae proceed, we put the sum of the
terms a, b, c, d, etc. equal to P, the sum of the squares equal to Q, the sum of the fourth
powers S, the sum of the fifth powers equal to T, the sum of the sixth powers equal to V,
etc. With these in place we have :
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 4
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

P   ; Q  P  2  ; R  Q  P  3 ; S  R  Q  P  4 ; T  S  R  Q  P  5 ; etc

§.9. Therefore when in our case 1A , p 1 A ,  p1 A , 2 p1 A ,  2 p1 A , 3 p1 A , 3 p1 A etc. ,

the sum of all the terms of the series or α is equal to 1y ; the sum of the factors from the
terms taken two at a time or β is equal to 0, and of those beyond :
  1.2.31. y ;   0;   1.2.3.41.5. y ; ς  0; etc The sum of these terms P  1y ; the sum of the
Q
squares of these terms Q  Py  12 ; the sum of the cubes of these terms R  y  1.21. y ; the
y
sum of the fourth powers S  Ry  1.2P.3 y . And thus again,
Q
T  Sy  1.2.3. y  [Link]. y ; V  Ty  1.2R.3. y  1.2.3P.4.5. y ;
Q
W  Vy  1.2S.3. y  [Link].5. y  [Link].5.6. y .
From which rule, the sums of the remaining higher powers can be easily determined.

§.10. Now we put the sine PM = y equal to the radius in order that y = 1, and the fourth
part of the circumference is the smallest arc A of this sine for which the sine is 1, equal to
1 p , or with q denoting the quarter of the circumference, then A = q and p = 2q. Hence
2
the above series is changed into this :
1 , 1 , 1 ,  1 , 1 ,  1 ,  1 ,  1 ,  1 ,  1 , etc.
q q 3q 3q 5q 5q 7q 7q 9q 9q
with two equal terms being present. Therefore the sum of these terms, which is :
2 (1  1  1  1  1  1  etc.) is equal to P which is equal to 1. Hence this series arises :
q 3 5 7 9 11
q p
(1  13  15  17  19  11
1  etc.)   . Therefore the fourth part of this series is equal to
2 4
the semi-circumference of the circle, of which the radius is 1, or to the whole
circumference of the circle of which the diameter is 1. And this is that series advanced
now some time ago by Leibniz , in which he defined the quadrature of the circle. From
which the power of this method, if perhaps one was not entirely sure of it, is manifestly
supported; thus as with the remaining series, which are to be derived by this method,
these are not to be doubted.

§.11. Now we take up the case of finding the terms when y = 1, the squares, and this
series is produced  12  12  1 2  1 2  1 2  1 2  etc. , the sum of which is
q q 9q 9q 25q 25q
2 (1  1  1  1  etc.) , which hence is equal to Q  P  1. From which it follows that
q2 1 9 25 49
1  1  etc. is equal to q  p ; with p denoting the whole
2 2
the sum of this series 1  19  25 49 2 8
circumference of the circle, the diameter of which is equal to 1. But the sum of this series
1  19  25
1  1  etc. depends on the sum of the series 1  1  1  1  1  etc . since this
49 4 9 16 25
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 5
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

series is a quarter part of that whole series. Hence the sum of this whole series is equal to
the sum of that series found and a third of it.
[As Q = sum odd squares + Q/4 or Q = 4/3  sum odd squares.]
p2
On account of which 1  14  19  16
1  1  1  etc. 
25 36 6
, and thus the sum of this series
multiplied by 6 is equal to the square of the circumference of the circle the diameter of
which is 1; which is that proposition that I mentioned in the beginning.

§.12. Therefore in the case when y = 1, on letting P = 1 and Q = 1, it follows that the
remaining letters R, S, T, V, etc., are as follows:
R  12 ; S  13 ; T  24
5 ; V  2 ; W  61 ; X  17 ; etc. But since for the sum of the cubes,
15 720 325
R  12 , then 23 (1  13  13  13  13 - etc.)  12 . Whereby the equation becomes :
q 3 5 7 9
q p3 3
1  13  13  13  13 - etc.  4  32 . Thus the sum of this series multiplied by 32 gives
3 5 7 9
the cube of the circumference of the circle of which the diameter is 1. In a similar manner
the sum of the fourth powers, which is 24 (1  14  14  14  14  etc.) must be equal to
q 3 5 7 9
1 , and thus 1  1  1  1  1  etc.  q  p . Now this series multiplied by 16 is
4 4

3 34 5 4 7 4 9 4 6 96 15
p 4

90
of that series; or to the series 1  14  14  14  14  14  etc. whereby this series is
2 3 4 5 6
p4
equal to 90 ; or for the 1  14  14  14  14  14  etc. , the sum multiplied by 90 gives
2 3 4 5 6
the fourth power of the circumference of the circle of which the diameter is 1.

§.13. In a similar way the sum of the higher powers can be found; moreover as it follows
5q 5 5 p5
that 1  15  15  15  15 - etc.  48  1536 ;
3 5 7 9
q p 6 6
then 1  16  16  16  16  etc.  16  960 .
3 5 7 9
Now from the sum of this series, likewise the sum of this series
p6
1  16  16  16  16  etc. becomes known, which is equal to 945 . Again for the seventh
2 3 4 5
61q 61 p 7 7
powers, there is 1  17  17  17  17 - etc.  1440  184320 and for the powers of eight
3 5 7 9
q8 17 p 8
1  18  18  18  18  etc.  630  161280 ; hence it is deduced that
3 5 7 9
p 8
1  18  18  18  18  18  etc.  9450 . Moreover it is to be observed that in these series
2 3 4 5 6
with the odd powers of the exponent with the alternating sign, that for the even powers
the signs are now the same; and because of this, since in the general
series, 1  1n  1n  1n  etc. the sum can only be shown in the cases in which n is an even
2 3 4
number. Besides also it is to be observed that if we find the terms of the series
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 6
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

5 , 2 , 61 , 17 etc. which are the values of the letters P, Q, R, S, etc., then the
1,1, 12 , 13 , 24 15 720 315
general sum can be assigned, while from that itself the quadrature of the circle can be
shown.

§.14. In these we have put the sine PM equal to the radius, hence we may see what series
are produced if other values are presented to y. Therefore, let y  1 , to which sine the
2
1 1
smallest corresponding arc is 4 p. Hence with A = 4 p , the series of the simplest terms
or of the first powers is thus 4p  34p  54p  74p  94p  114p  etc. , and the sum P of this

series is equal to 1y  2 . Hence there is obtained :

p
 1  13  15  17  19  11
1  1  1  etc. , which series differs from that of Leibniz only by
13 15
2 2
reason of the sign, and advanced by Newton now a long time ago. Now the sum of the
1  1  etc.) is itself equal to Q  2 . Hence
squares of these terms, clearly 162 (1  19  25
p 49
1  1  etc.  p , as was found before.
2
there is 1  19  25 49 8

§.15. If we make y  3 , the smallest arc of this sine corresponds to 600, and thus
2
1
A = 3 p . Therefore in this case the series of terms is produced :
3  3  3  3  3  3  etc. , and the sum of which terms is equal to 1  2 .
p 2p 4p 5p 7p 8p y 3
2p
Therefore it is found that  1  12  14  15  17  18  10
1  1  etc. Now the sum of the
11
3 3
squares of these terms is equal to 2  3 ; thus it follows to be the case that
1 4
y
4 p2
27
 1  14  16
1  1  1  1  etc. in which series the successive third terms are
25 49 64
missing. But this series also depends on that series 1  14  92  16
3 etc. , the sum of which

p2
was found to be equal to 6 ; for if this series is diminished by its ninth part then the
p2 4 pp
above series is produced, thus the sum of this must be equal to 6 (1  19 )  27 . In a
similar manner, if other sines are assumed, then other series are produced, as of the first
power and then of the squares and of higher powers, and the sums of which involve the
quadrature of the circle.
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 7
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

§.16. But if y is put equal to 0, it is not possible to assign series of this kind any more. on
account of placing y in the denominator, or the initial series divided by y. But the series
are able to be deduced in another way, while which series are of the form
1  1n  1n  1n  etc. , if n is an even number : as the sums of these series are found, I
2 3 4
can deduce separately from these the case in which y = 0. Now with y = 0 the
fundamental equation is turned into this equation :
0  s  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc. ,
3 5 7

and the roots of this equation give all the arcs, of which the sine is equal to zero. But the
smallest root is s = 0, whereby the equation divided by s shows all the remaining arcs of
which the sine is equal to zero, which arcs are hence produced by the roots of this
equation : 0  1  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc.
2 4 6

Now the arcs of these of which the sine is equal to zero are :
p ,  p , 2 p ,  2 p , 3 p ,  3 p etc . of which each second term is negative, that which the
equation also indicates on account of only having even dimensions of s. Whereby the
divisors of this equation are :
1  sp ,1  sp ,1  2sp ,1  2sp , etc. ,
and with the pairs of these divisors joined together, there is the equation :
1  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc.  (1 - s 2 )(1 - s 2 )(1 - 5s 2 )(1 - s 2 ) etc.
2 4 6 2 2 2 2

p 4p 9p 16 p

§.17. Now it is evident from the nature of the equations, that the coefficient of ss or 1.21.3
is equal to 12  1 2  1 2  1 2  etc. Now the sum of the factors from the two terms of
p 4p
9p 16 p
this series is equal to [Link].5 ; and the sum of the factors from the three terms is equal to
1 etc. On account of this, just as in §.8
[Link].5.6.7
  1.21.3 ;   [Link].5 ;   [Link].5.6.7 ; etc. and with the sum of the terms also put equal
to P : 12  1 2  1 2  1 2  etc.  P, and with the sum of the squares of the terms
p 4p 9p 16 p
equal to Q; the sum of the cubes equal to R; the sum of the fourth powers equal to S, etc. ,
by §.8 we have : P    1.21.3  16 ; Q  P  2   90
1 ; R  Q  P  3  1 ;
945
S  R  Q  P  4  9450
1 T  S  R  Q  P  5  1 ;
93555
V  T  S  R  Q  P  6  6825691
 93555
etc.
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 8
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

§.18. Hence from these the following sums can be derived :

p2
1  12  12  12  12  etc.  6  P
2 3 4 5
p4
1  14  14  14  14  etc.  90  Q
2 3 4 5
p 6
1  16  16  16  16  etc.  945  R
2 3 4 5
p8
1  18  18  18  18  etc.  9450  S
2 3 4 5
p 10
1  110  110  110  110  etc.  93555  T
2 3 4 5
691 p 12
1  112  112  112  112  etc.  682193555  V .
2 3 4 5

which series from the given law but yet with much labour can be extended to the higher
powers. Moreover, with a particular series divided by the preceding series, this sequence
of equations arises :
15Q R  10 S  99T  6825V etc. from which individual expressions the
p 2  6 P  P  21
2Q R 10 S 691T
square of the perimeter of the circle of which the diameter is 1 is equal.
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 9
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

§.19. Moreover when the sums of these series, even if now they can be easily shown as
above, yet they are not able to bring much help to expressing the perimeter of the circle
on account of the root of the square, that has to be extracted; from the former series we
can elicit expressions, which are equal to the periphery p itself . Moreover, these are
produced as follows :

p  4(1  13  15  17  19  11
1  etc.)

 1 312  512  712  912  1112 etc. 

p  2 
1 13  15  17  19  111  etc. 
 

 1 313  513  713  913 - 1113 etc. 

p  4 1 1 1 1 1 
1 2  2  2  2  2  etc. 
 3 5 7 9 11 

 1 314  514  714  914  1114 etc. 

p  8 1 1 1 1 1 
1 3  3  3  3 - 3 etc. 
 3 5 7 9 11 

 1 315  515  715  915 -1115 etc. 

p   1 1 1 1 1
16 
5 1 4  4  4  4  4  etc. 
 3 5 7 9 11 

 1 316  516  716  916  1116 etc. 

p  8  1 1 1 1 1
25 
1 5  5  5  5 - 5 etc. 
 3 5 7 9 11 

 1 317  517  717  917 - 1117 etc. 

p  61  1 1 1 1 1
192 
1 6  6  6  6  6  etc.  .
 3 5 7 9 11 
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 10
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

De summis serierum reciprocarum.

Auctore Leonh. Eulero.

§. 1.
Tantopere iam pertractatae et investigatae sunt series reciprocae potestatum
numerorum naturalium, ut vix probabile videatur de iis novi quicquam invenire posse.
Quicunque enim de summis serierum meditati sunt, ii fere omnes quoque in summas
huiusmodi serierum inquisiverunt, neque tamen ulla methodo eas idoneo modo exprimere
potuerunt. Ego etiam iam saepius, cum varias summandi methodos tradidissem, has series
diligenter sum persecutus, neque tamen quicquam aliud sum assecutus, nisi ut earum
summam vel proxime veram definiverim vel ad quadraturas curvarum maxime
transcendentium reduxerim; quorum illud in differtatione proxime praelecta, hoc vero in
praecedentibus praestiti. Loquor hic autem de seriebus fractionum, quarum numeratores
sunt 1, denominatores vero vel quadrata, vel cubi, vel aliae dignitates numerorum
naturalem : cuius modi sunt 1  14  19  16
1  1  etc. , item 1  1  1  1  etc. atque
25 8 27 64
similes superiorum potestatum, quarum termini generales continentur in hac forma 1n .
x
§.2. Deductus sum autem nuper omnino inopinato ad elegantem sumae huius seriei
1  14  19  16
1  1  etc. expressionem, quae a circuli quadratura pendet, ita, ut si hius
25
seriei vera summa habetetur, inde simul circuli quadratura sequeratur. Inveni enim
summae huius seriei sextuplum aequale esse quadrato peripheriae circuli, cuius diameter
est 1; seu posita istius seriei summa = s, tenebit 6s ad 1 rationem peripheriae ad
diametrum. Huius autem seriei summam nuper ostendi proxime esse 1,
6449340668482264364, ex cuius numeri sextuplo, si extrahatur radix quadrata, reipsa
prodit numerus 3,141592653589793238 exprimens circuli peripheriam, cuius diameter
est 1. Iisdam porro vestigiis quibus hanc summam sum consecutus, incedens huius seriei
1  16
1  1  1  1  etc. summam quoque a quadratura circuli pendere deprehendi.
81 256 625
Summa nempe eius per 90 multiplicata dat biquadratum peripheriae circuli, cuius
diameter est 1. Atque simili ratione etiam sequentium serierum, in quibus exponentes
dignitatum sunt numeres pares, summas determinare potui.

§.3. Quo igitur, quemadmodum haec sum adeptus,

commodissime ostendam, totam rem, quo ipse usus
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 11
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

sum, ordine exponam. In circulo AMBNA centro C radio AC vel BC = 1 descripto

contemplatus sum arcum quemcunque AM, cuius sinus est MP, cosinus vero CP. Posito
nunc arc AM = s, sinu PM = y, et cosinu CP = x, per methodum iam satis cognitam tam
sinus y cosinusx ex dato arcu s per series possunt definiri, est enim, uti passim videre licet
y  s  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc. atque x  1  1s.2  1.2s.3.4  1.2.3s.4.5.6  etc. Ex
3 5 7 2 4 6

harum scilicit aequationem consideratione ad summas supra memoratarum serierum

reciprocarum perveni; quarum aequationum quidem utraque ad eundem fere scopum
dirigitur, et hanc ob rem sufficiet alteram tantum eo, quem sum expositurus, modo
tractasse.

§.4. Aequatio ergo prior y  s  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc. exprimit relationem
3 5 7

inter arcum et sinum. Quare ex ea tam ex dato arcu eius sinus, quam ex dato sinu eius
arcus determinari poterit. Considero autem sinum y tanquam datum, et investigo,
quemadmodum arcum s ex y erui oporteat. Hic vero ante omnia animadvertendum est,
eidem sinui y innumerabiles arcus respondere, quos ergo innumerabiles arcus aequatio
proposita praebere debebit. Si quidem in ista aequatio s tanquam incognita spectetur, ea
infinitas habet dimensiones, ideoque mirum non est, si ista aequatio innumeros contineat
factores simplices, quorum quisque nihilo aequalis positus, idoneum pro s valorem dare
debet.

§.5. Quemadmodum autem, si omnes factores huius aequationis cogniti essent, omnes
quoque radices illius seu valores cogniti essent, omnes quoque radices illius valores s
innotescerent, ita vicissim si omnes valores ipsius s assingari poterunt, tum quoque ipsi
factores omnes habebuntur. Quo autem eo melius tam de radicibus quam de factoribus
iudicare queam, transmuto aequationem propositam in hanc formam :
0  1  sy  1.2s.3. y  1.2.3s.4.5. y  etc. Si nunc omnes radices huius aequationis seu omnes
3 5

arcus, quorum idem est sinus y, fuerint A, B, C, D, E, etc. tum factores quoque erunt
omnes istae quantitates, 1  As , 1  Bs , 1  Cs , 1  Ds , etc. Quamobrem erit

0  1  sy  1.2s.3. y  1.2.3s.4.5. y  etc.  ( 1  As ) ( 1  Bs ) ( 1  Cs ) ( 1  Ds ) etc.

3 5

§.6. Ex natura autem et resolutione aequationum constat, esse coefficentem termini, in

quo inest s, seu 1y aequalem summae omnium coefficientium ipsius s in factoribus seu
1
y
 1A  B1  C1  D 1  etc. Deinde est coefficiens ipsius s2, qui est = 0, ob hunc terminum

in aequatione deficientem, aequalis aggregatio factorum ex binis terminis seriei,

1 , 1 , 1 , 1 , etc. Porro erit  1 aequale aggregato factorum ex quaternis terminis
A B C D 1.2.3. y
eiusdem seriei 1A , B1 , C1 , D
1 , etc. Similique modo erit 0 = aggregato factorum ex quaternis

terminis eiusdem seriei, et  [Link].5. y  aggregato factorum ex quinis terminis istius

seriei, et ita porro.
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 12
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

§.7. Posito autem mimimo arcu AM = A, cuius sinus est PM = y, et semiperipheria circuli
= p, erunt A, p  A, 2 p  A, 3 p  A, 4 p  A, 5 p  A, 6 p  A etc.
item  p  A,  2 p  A,  3 p  A,  4 p  A,  5 p  A, etc. omnes arcus, quorum sinus est
idem y. Quam igitur ante assumsimus seriem 1A , B1 , C1 , D 1 , etc. ea transmutatur in

hanc 1A , p 1 A ,  p1 A , 2 p1 A ,  2 p1 A , 3 p1 A , 3 p1 A , 4 p1 A ,  4 p1 A etc. Horum ergo omnium

terminorum summa est = 1y ; summa autem factorum ex binis terminis huius seriei est

aequalis 0; summa factorum ex ternis = 1.2.31. y , summa factorum ex quaternis = 0;

summa factorum ex quinis = 1.2.3.41.5. y ; summa factorum ex senis = 0. Atque ita porro.

§.8. Si autem habeatur series quaecunque a  b  c  d  e  f  etc. cuius summa sit α,

summa factorum ex binis terminis = β; summa factorum ex ternis = γ; summa factorum
ex quaternis = δ, etc. erit summa quadratorum singulorum terminorum, hoc est
a 2  b 2  c 2  d 2  e 2  f 2  etc. =  2  2 ; summa vero cuborum
a3  b3  c3  d 3  e3  f 3  etc. =  3  3  3 ; summa biquadratorum =
 4  4 2   4  2 2  4 . Quo autem clarius appareat, quomodo hae formulae
progrediantur, ponamus ipsorum terminorum a, b, c, d, etc. summam esse = P, summam
quadratorum = Q, summa biquadratorum = S, summam potestatum quintarum = T,
summam sextarum = V etc. His positis erit
P   ; Q  P  2 ; R  Q  P  3 ; S  R  Q  P  4 ; T  S  R  Q  P  5 ; etc

§.9. Cum igitur in nostro casu seriei

1, 1 , 1 , 1 , 1 , 1 , 1 etc. summa omnium terminorum seu α sit
A p  A  p  A 2 p  A  2 p  A 3 p  A 3 p  A
= 1y ; summa factorum ex binis seu β = 0, atque ulterius

  1.2.31. y ;   0;   1.2.3.41.5. y ; ς  0; etc erit summa ipsorum illorum terminorum

P  1y ; summa quadratorum illorum terminorum Q  Py  12 ; summa cuborum illorum
y
Q
terminorum R  y  1.21. y ;summa biquadratorum S  Ry  1.2P.3 y . Atque porro
Q
T  Sy  1.2.3. y  [Link]. y ; V  Ty  1.2R.3. y  1.2.3P.4.5. y ;
Q
W  Vy  1.2S.3. y  [Link].5. y  [Link].5.6. y . Ex qua lege facile reliquarum altiorum
potestatum summae determinantur.
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 13
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

§.10. Ponamus nunc sinum PM = y aequalem radio, ut sit y = 1, erit minimus arcus A
cuius sinus est 1 quarta peripheria pars, = 12 p , seu denotante q quartam peripheriae
partem erit A = q et p = 2q. Superior ergo series abibit in istam
1 , 1 , 1 ,  1 , 1 ,  1 ,  1 ,  1 ,  1 ,  1 , etc. binis terminis existentibus
q q 3q 3q 5q 5q 7q 7q 9q 9q
aequalibus. Horum ergo terminorum summa, quae est
2 (1  1  1  1  1  1  etc.) aequalis est ipsi P = 1. Hinc igitur oritur
q 3 5 7 9 11
q p
(1  13  15  17  19  11
1  etc.)   . Huius ergo seriei quadruplum aequatur
2 4
semiperipheriae circuli, cuius radius est 1, seu toti peripheriae circuli, cuius diameter est
1. Atque haec est ipsa series a Leibnitio iam pridem prolata, qua circuli quadraturam
definiuit. Ex quo magnum huius methodi, si cui forte ea non satis certa videatur,
firmamentum elucit; ita ut de reliquis, quae ex hac methodo derivabantur, omnino non
liceat dubitari.

§.11. Sumamus nunc inventorum terminorum pro casu quo y = 1, quadrata, prodibitque
haec series  12  12  1 2  1 2  1 2  1 2  etc. cuius summa est
q q 9q 9q 25q 25q
2 (1  1  1  1  etc.) , quae ergo aequalis esse debet ipsi Q  P  1. Ex quo sequitur
q 2 1 9 25 49
q2 p2
huius seriei 1  19  25
1  1  etc. summam esse =
49 2
 8 ; denotante p totam circuli
peripheriam, cuius diameter est = 1. Summa autem huius seriei 1  19  25
1  1  etc.
49
pendet a summa seriei 1  14  19  16
1  1  etc. quia haec quarta sui parte minuta illam
25
dat. Est ergo summa huius seriei aequalis summa illius sui triente. Quamobrem erit
p2
1  14  19  16
1  1  1  etc. 
25 36 6
, ideoque huius seriei summa per 6 multiplicata
aequalis est quadrato peripheriae circuli cuius diameter est 1; quae est ipsa propositio
cuius initio mentionem feci.

§.12. Cum igitur casu quo y = 1, sit P = 1 et Q = 1, erunt reliquarum litterarum R, S, T, V,

etc. ut sequitur : R  12 ; S  13 ; T  24
5 ; V  2 ; W  61 ; X  17 ; etc. Cum autem
15 720 325
summa cuborum ipsi R  12 sit aequalis, erit 23 (1  13  13  13  13 - etc.)  12 . Quare erit
q 3 5 7 9
q
3
p 3
1  13  13  13  13 - etc.  4  32 . Huius ideo seriei summa per 32 multiplicat dat
3 5 7 9
cubum peripheriae circuli cuius diameter est 1. Simili modo summa biquadratorum, quae
est 24 (1  14  14  14  14  etc.) aequalis esse debet 13 , ideoque erit
q 3 5 7 9
q4 p4
1  14  14  14  14  etc.  6  96 . Est vero haec series per 16
15
multiplicata ista series
3 5 7 9
p 4
aequalis est 90 ; seu seriei 1  14  14  14  14  14  etc. quare ista series aequalis est
2 3 4 5 6
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 14
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

p4
90
; seu seriei 1  14  14  14  14  14  etc. summa per 90 multiplicata dat biquadratum
2 3 4 5 6
peripheriae circuli cuius diameter est 1.

§.13. Simili modo invenientur summae superiorum potestatum; prodibit autem ut sequitur
5q 5p 5 5
1  15  15  15  15 - etc.  48  1536 ; atque
3 5 7 9
q p 6 6
1  16  16  16  16  etc.  16  960 . Inventa vero huius seriei summa, cognoscetur simul
3 5 7 9
p6
summa huius seriei 1  16  16  16  16  etc. quae erit = 945 . Porro pro potestatibus
2 3 4 5
61q 61 p 7 7
septimus erit 1  17  17  17  17 - etc.  1440  184320 ac pro octavis
3 5 7 9
q8 17 p 8
1  18  18  18  18  etc.  630  161280 ; unde deducitur
3 5 7 9
p 8
1  18  18  18  18  18  etc.  9450 . Observandum autem est de his seriebus in
2 3 4 5 6
potentiis exponentium imparium signa terminoru alternari, pro potestatibus paribus vero
esse aequalia; hocque in cause est, quod huius generalis seriei 1  1n  1n  1n  etc. iis
2 3 4
tantum casibus summa possit exhiberi, quibus n est numerus par. Praeterea quoque
5 , 2 , 61 , 17 etc. quos valores pro litteris P, Q, R, S,
notandum est, si seriei 1,1, 12 , 13 , 24 15 720 315
etc. invenimus, terminus generalis posset assignari, tum eo ipso quadraturam circuli
exhibitum iri.

§.14. In his posuimus sinum PM aequalem radio, videamus ergo quales series prodeant, si
ipsi y alii valores tribuantur. Sit igitur y  1 , cui sinui minimus arcus respondens est
2
1 p . Posito ergo A = 1 p erit series terminorum simplicium seu primae potestatis ista
4 4
4  4  4  4  4  4  etc. cuius seriei summa P aequalis est 1  2 . Habibitur
p 3 p 5 p 7 p 9 p 11 p y
p
ergo  1  13  15  17  19  11
1  1  1 etc. quae series tantum ratione signorum a
13 15
2 2
Leibnitiana differt, et a Newtono iam dudum est prolata. Summa vero quadratorum
illorum terminorum nempe 162 (1  19  25
1  1  etc.) aequalis est ipsi Q = 2. Erit ergo
p 49
1  1  etc.  p , uti ante est inventum.
2
1  19  25 49 8

§.15. Si fiat y  3 erit minimus arcus huic sinui respondens 600, ideoque A = 13 p . Hoc
2
ergo casu sequens prodibit series terminorum 3p  23p  43p  53p  73p  83p  etc. quorum

terminorum summa aequalis est ipsi 1y  2 . Habitur ergo

3
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 15
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

2p
 1  12  14  15  17  18  10
1  1  etc. Summa vero quadratorum illorum terminorum
11
3 3
4p 2
est = 12  34 ; unde sequitur fore 27  1  14  16
1  1  1  1  etc. in qua serie desunt
y 25 49 64
termini ternario constantes. Pendet autem haec series quoque ab ista
p2
1  14  92  16
3 etc. cuius summa erat inventa =
6
; nam si haec series sui parte nona
p2 4 pp
minuatur prodit ipsa superior series, cuius ideo summa debet esse = 6 (1  19 )  27 .
Simili modo si alii assumatur sinus, aliae prodibunt series, tam simplicium, quam
terminorum quadratorum altiorumque potestatum, quarum summae quadraturum circuli
involuent.

§.16. At si ponatur y = 0, huiusmodi series non amplius assignari poterunt, propter y in

denominatorem positum, seu aequationem inialem per y divisam. Alio autem modo series
inde deduci poterunt, quae cum sint ipsae series 1  1n  1n  1n  etc. si n est numerus
2 3 4
par : quaemadmodum harum serierum summae sint inveniendae, seorsum ex hoc casu
quo y = 0 deducam. Posito vero y = 0 ipsa aequatio fundamentalis abit in hanc
0  s  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc. cuius aequationis radices dant omnes arcus,
3 5 7

quorum sinus est = 0. Est autem una minimaque radix s = 0, quare aequatio per s divisa
exhibebit reliquos arcus omnes, quorum sinus est = 0, qui arcus proinde erunt radices
huius aequationis 0  1  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc. Ipsi vero arcus quorum sinus
2 4 6

est = 0 sunt p ,  p , 2 p ,  2 p , 3 p ,  3 p etc . quorum binorum alter alterius est negativus, id

quod quoque ipsa aequatio propter dimensiones ipsius s tantum pares indicat. Quare
divisores illius aequationis erunt 1  sp ,1  sp ,1  2sp ,1  2sp , etc. atque coniungendis binis
horum divisorum erit
1  1.s2.3  1.2.s3.4.5  1.2.3.s4.5.6.7  etc.  (1 - s 2 )(1 - s 2 )(1 - 5s 2 )(1 - s 2 ) etc.
2 4 6 2 2 2 2

p 4p 9p 16 p

§.17. Manifestum iam est ex natura aequationum, fore coefficientem ipsius ss seu 1.21.3
aequalem 12  1 2  1 2  1 2  etc. Summa vero factorum ex binis terminis huius
p 4p 9p 16 p
seriei erit  [Link].5 ; summaque factorum ex ternis  [Link].5.6.7 etc. Hanc ob rem erit
iuxta §.8   1.21.3 ;   [Link].5 ;   [Link].5.6.7 ; etc. atque posita quoque summa
terminorum 12  1 2  1 2  1 2  etc.  P, et summa quadratorum eorundem
p 4p 9p 16 p
terminorum = Q; summa cuborum = R; summa biquadratorum = S, etc. , erit per §.8
P    1.21.3  16 ; Q  P  2   90
1 ; R  Q  P  3  1 ;
945
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 16
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

S  R  Q  P  4  9450
1 T  S  R  Q  P  5  1 ;
93555
V  T  S  R  Q  P  6  6825691
 93555
etc.

§.18. Ex his ergo derivantur summae sequentes :

p2
1  12  12  12  12  etc.  6  P
2 3 4 5
p4
1  14  14  14  14  etc.  90  Q
2 3 4 5
p 6
1  16  16  16  16  etc.  945  R
2 3 4 5
p8
1  18  18  18  18  etc.  9450  S
2 3 4 5
p 10
1  110  110  110  110  etc.  93555  T
2 3 4 5
691 p 12
1  112  112  112  112  etc.  682193555  V .
2 3 4 5

quae series ex data lege attamen multo laboread altiores potestates produci possunt.
Dividendis autem singulis seriebus per praecedentes orientur sequentes aequationes :
15Q R  10 S  99T  6825V etc. quibus expressionibus singulis quadratum
p 2  6 P  P  21
2Q R 10 S 691T
peripheriae cuius diameter est 1, aequatur.

§.19. Cum autem harum serierum summae etiamsi vero proxime facile exhiberi possent,
tamen non multum adiumenti afferre queant ad peripheriam circuli vero proxime
exprimendam propter radicem quadratam, quae extrahi deberet; ex prioribus seriebus
elicemus expressiones, quae ipsi peripheriae p sint aequales. Prodibit autem vi sequitur :

p  4(1  13  15  17  19  11
1  etc.)

 1 312  512  712  912  1112 etc. 

p  2 
1 13  15  17  19  111  etc. 
 

 1 313  513  713  913 - 1113 etc. 

p  4 1 1 1 1 1 
1 2  2  2  2  2  etc. 
 3 5 7 9 11 

 1 314  514  714  914  1114 etc. 

p  8 1 1 1 1 1 
1 3  3  3  3 - 3 etc. 
 3 5 7 9 11 
 Comment. acad. sc. Petrop. 7. 1740 (p. 124 onwards) : 17
L. EULER. De summis serierum reciprocarum. (E041). Translated by Ian Bruce.

 1 315  515  715  915 -1115 etc. 

p   1 1 1 1 1
16 
5 1 4  4  4  4  4  etc. 
 3 5 7 9 11 

 1 316  516  716  916  1116 etc. 

p  8  1 1 1 1 1
25 
1 5  5  5  5 - 5 etc. 
 3 5 7 9 11 

 1 317  517  717  917 - 1117 etc. 

p  61  1 1 1 1 1
192 
1 6  6  6  6  6  etc.  .
 3 5 7 9 11 