Each inverse trigonometric function is the inverse of a trigonometric function restricted to an
interval on which that function is one-one, and its derivative can be found by writing the
inverse relation as an ordinary (implicit) equation and differentiating both sides -- exactly the
technique Section 5 develops in general, applied here to the six standard inverse functions.
Derivative of sin−1x. Let y=sin−1x, for x∈(−1,1) and y∈(−π/2,π/2);
this means siny=x. Differentiating both sides with respect to x (using the chain rule on
the left, since y is a function of x):
cosy⋅dxdy=1⟹dxdy=cosy1.
Since y∈(−π/2,π/2), cosy>0, so cosy=1−sin2y=1−x2 (taking the
positive root). Hence
dxd(sin−1x)=1−x21,−1<x<1.
Derivative of cos−1x. By an identical argument with y=cos−1x (cosy=x,
y∈(0,π), so siny>0 and siny=1−x2): differentiating cosy=x gives
−siny(dy/dx)=1, so
dxd(cos−1x)=−1−x21,−1<x<1.
(Note sin−1x+cos−1x=π/2 for all x∈[−1,1] -- a constant -- so their derivatives
must be negatives of each other, exactly as found; this is a useful self-check.)
Derivative of tan−1x. Let y=tan−1x, so tany=x for all real x, with
y∈(−π/2,π/2). Differentiating: sec2y(dy/dx)=1, and using sec2y=1+tan2y=1+x2,
dxd(tan−1x)=1+x21,x∈R.
Derivative of cot−1x. By the same method with coty=x: −cosec2y(dy/dx)=1,
and cosec2y=1+cot2y=1+x2, so
dxd(cot−1x)=−1+x21,x∈R.
Derivatives of sec−1x and cosec−1x (for ∣x∣>1), obtained the same way from
secy=x and cosecy=x respectively:
dxd(sec−1x)=∣x∣x2−11,dxd(cosec−1x)=−∣x∣x2−11.
Combined with the chain rule. For a composite argument, e.g. y=sin−1(x2), treat u=x2 …