TOPICS
Search

Lemniscate Function

DOWNLOAD Mathematica NotebookDownload Wolfram Notebook

The lemniscate functions arise in rectifying the arc length of the lemniscate. The lemniscate functions were first studied by Jakob Bernoulli and Giulio Fagnano. A historical account is given by Ayoub (1984), and an extensive discussion by Siegel (1969). The lemniscate functions were the first functions defined by inversion of an integral

L=2aint_0^1(1-t^4)^(-1/2)dt,
(1)

which was first done by Gauss, who noticed that

1/(agm(1,sqrt(2)))=2/piint_0^1(dt)/(sqrt(1-t^4)),
(2)

where agm(a,b) is the arithmetic-geometric mean (Borwein and Bailey 2003, p. 13).

Define the inverse lemniscate functions as

phi(x)=sinlemn^(-1)x
(3)
=int_0^x(1-t^4)^(-1/2)dt
(4)
=x_2F_1(1/4,1/2;5/4;x^4)
(5)
phi^'(x)=coslemn^(-1)x
(6)
=int_x^1(1-t^4)^(-1/2)dt
(7)
=K(i)-F(sin^(-1)x,i)
(8)
=(Gamma^2(1/4))/(4sqrt(2pi))-x_2F_1(1/4,1/2;5/4;x^4),
(9)

where _2F_1(a,b;c;z) is a hypergeometric function, F(z,k) is an incomplete elliptic integral of the first kind, K(k) is an elliptic integral of the second kind, and

pi=L/a,
(10)

so that

x=sinlemnphi
(11)
x=coslemnphi^'.
(12)

Now, there is an identity connecting phi and phi^' since

phi(x)+phi^'(x)=L/(2a)=1/2pi,
(13)

so

sinlemnphi=coslemn(1/2pi-phi).
(14)

These functions can be written in terms of Jacobi elliptic functions,

u=int_0^(sd(u,k))[(1-k^('2)y^2)(1+k^2y^2)]^(-1/2)dy.
(15)

Now, if k=k^'=1/sqrt(2), then

u=int_0^(sd(u,1/sqrt(2)))[(1-1/2y^2)(1+1/2y^2)]^(-1/2)dy
(16)
=int_0^(sd(u,1/sqrt(2)))(1-1/4y^4)^(-1/2)dy.
(17)

Let t=y/sqrt(2) so dy=sqrt(2)dt,

u=sqrt(2)int_0^(sd(u,1/sqrt(2))/sqrt(2))(1-t^4)^(-1/2)dt
(18)
u/(sqrt(2))=int_0^(sd(u,1/sqrt(2))/sqrt(2))(1-t^4)^(-1/2)dt
(19)
u=int_0^(sd(usqrt(2),1/sqrt(2))/sqrt(2))(1-t^4)^(-1/2)dt,
(20)

and

sinlemnphi=1/(sqrt(2))sd(phisqrt(2),1/(sqrt(2))).
(21)

Similarly,

u=int_(cn(u,k))^1(1-t^2)^(-1/2)(k^('2)+k^2t^2)^(-1/2)dt
(22)
=int_(cn(u,1/sqrt(2)))^1(1-t^2)^(-1/2)(1/2+1/2t^2)^(-1/2)dt
(23)
=sqrt(2)int_(cn(u,1/sqrt(2)))^1(1-t^4)^(-1/2)dt
(24)
u/(sqrt(2))=int_(cn(u,1/sqrt(2)))^1(1-t^4)^(-1/2)dt
(25)
u=int_(cn(usqrt(2),1/sqrt(2)))^1(1-t^4)^(-1/2)dt,
(26)

and

coslemnphi=cn(phisqrt(2),1/(sqrt(2))).
(27)

We know

coslemn(1/2pi)=cn(1/2pisqrt(2),1/(sqrt(2)))=0.
(28)

But it is true that

cn(K,k)=0,
(29)

so

K(1/(sqrt(2)))=1/2sqrt(2)pi=1/(sqrt(2))pi
(30)
(Gamma^2(1/4))/(4sqrt(pi))=1/(sqrt(2))pi
(31)
L=api=(Gamma^2(1/4))/(2^(3/2)sqrt(pi))a.
(32)

By expanding (1-t^4)^(-1/2) in a binomial series and integrating term by term, the arcsinlemn function can be written

phi(x)=int_0^x(dt)/(sqrt(1-t^4))
(33)
=sum_(n=0)^(infty)((1/2)_nx^(4n+1))/(n!(4n+1))
(34)
=x_2F_1(1/4,1/2;5/4;x^4),
(35)

where (a)_n is a Pochhammer symbol (Berndt 1994).

Ramanujan gave the following inversion formula for phi(x). If

(thetamu)/(sqrt(2))=sum_(n=0)^infty((1/2)_nx^(4n+1))/(n!(4n+1)),
(36)

where

mu=(Gamma^2(1/4))/(2pi^(3/2))
(37)

is the constant obtained by letting x=1 and theta=pi/2, and

v=2^(-1/2)sd(mutheta),
(38)

then

(mu^2)/(2x^2)=csc^2theta-1/pi-8sum_(n=1)^infty(ncos(2ntheta))/(e^(2pin)-1)
(39)

(Berndt 1994).

Ramanujan also showed that if 0<theta<pi/2, then

-mu/(sqrt(2))sum_(n=0)^infty((1/2)_nv^(4n-1))/(n!(4n-1))=cottheta+theta/pi+4sum_(n=1)^infty(sin(2ntheta))/(2^(2pin)-1),
(40)
lnv+1/6pi-1/2ln2+sum_(n=1)^infty((1/4)_nv^(4n))/((3/4)_n4n)=ln(sintheta)+(theta^2)/(2pi)-2sum_(n=1)^infty(cos(2ntheta))/(n(e^(2pin)-1)),
(41)
1/2tan^(-1)v=sum_(n=0)^infty(sin[(2n+1)theta])/((2n+1)cosh[1/2(2n+1)pi])
(42)
1/4cos^(-1)(v^2)=sum_(n=0)^infty((-1)^ncos[(2n+1)theta])/((2n+1)cosh[1/2(2n+1)pi]),
(43)

and

(sqrt(2))/(4mu)sum_(n=0)^infty(2^(2n)(n!)^2)/((2n+1)!(4n+3))v^(4n+3)=(pitheta)/8-sum_(n=0)^infty((-1)^nsin[(2n+1)theta])/((2n+1)^2cosh[1/2(2n+1)pi])
(44)

(Berndt 1994).

A generalized version of the lemniscate function can be defined by letting 0<=theta<=pi/2 and 0<=v<=1. Write

2/3thetamu=int_0^v(dt)/(sqrt(1-t^6)),
(45)

where mu is the constant obtained by setting theta=pi/2 and v=1. Then

mu=(sqrt(pi))/(Gamma(2/3)Gamma(5/6)),
(46)

and Ramanujan showed

(4mu^2)/(9v^2)=csc^2theta-2/(pisqrt(3))+8sum_(n=1)^infty((-1)^(n-1)ncos(2ntheta))/(e^(pinsqrt(3))-(-1)^n)
(47)

(Berndt 1994).

See also

Arithmetic-Geometric Mean, Elliptic Function, Elliptic Integral, Gauss's Constant, Hyperbolic Lemniscate Function

Explore with Wolfram|Alpha

References

Ayoub, R. "The Lemniscate and Fagnano's Contributions to Elliptic Integrals." Arch. Hist. Exact Sci. 29, 131-149, 1984.Berndt, B. C. Ramanujan's Notebooks, Part IV. New York: Springer-Verlag, pp. 245, and 247-255, 258-260, 1994.Borwein, J. and Bailey, D. Mathematics by Experiment: Plausible Reasoning in the 21st Century. Wellesley, MA: A K Peters, 2003.Siegel, C. L. Topics in Complex Function Theory, Vol. 1. New York: Wiley, 1969.

Referenced on Wolfram|Alpha

Lemniscate Function

Cite this as:

Weisstein, Eric W. "Lemniscate Function." From MathWorld--A Wolfram Resource. https://mathworld.wolfram.com/LemniscateFunction.html

Subject classifications