Certain limits recur so often — and cannot be found by mere substitution — that their results are memorised as standard limits. Each is stated below with the exact conditions under which it holds.
Algebraic standard limit. For any rational index n and any a for which the expressions are defined,
limx→ax−axn−an=nan−1.
This follows from factoring xn−an=(x−a)(xn−1+xn−2a+⋯+an−1) and cancelling (x−a); the bracket has n terms, each tending to an−1, giving nan−1.
Trigonometric standard limits (angle x measured in radians):
limx→0xsinx=1,limx→0xtanx=1,limx→0x21−cosx=21.
Exponential and logarithmic standard limits:
limx→0xax−1=logea (a>0),limx→0xex−1=1,
limx→0xloge(1+x)=1,limx→0(1+x)1/x=e.
The result limx→0xex−1=1 is just the case a=e of the first (since logee=1).
How to use them — match the pattern. Each standard limit fits a shape. To apply one, algebraically force the expression into that exact shape, adjusting by a constant factor if the argument is a multiple of x. For example, xsin5x=5⋅5xsin5x, and as x→0 the factor 5xsin5x→1 (since 5x→0), so the whole limit is 5⋅1=5.
Radians, and the Argument Must Match the Denominator …