Closing in on 2
Start at 1 and go halfway to 2: 1.5. Go halfway again: 1.75. Then 1.875, then 1.9375. The gap to 2 is 1, then 0.5, 0.25, 0.125, 0.0625, halving every time.
No step lands on 2, because half of a gap is never zero. But the values get as close to 2 as you like: after ten steps the gap is , less than 0.001. That is what it means to say the values approach 2, and 2 is their limit.
The values 1, 1.5, 1.75, 1.875 and 1.9375, each halfway from the one before to 2. They crowd toward the gold mark at 2 and never reach it.
Near 2, not at 2
For a function, the question is what f(x) approaches as x approaches a number c. Saying that the limit of f(x) as is L means that f(x) can be made as close to L as you like by taking x close enough to c, from either side, with .
Take f(x) = x + 1 and c = 2. At x = 1.9, 1.99 and 1.999, f(x) is 2.9, 2.99 and 2.999. At x = 2.1, 2.01 and 2.001 it is 3.1, 3.01 and 3.001. From both sides the outputs close in on 3, so the limit of x + 1 as is 3.
Here f(2) = 3 as well, but the limit did not use it. Only inputs near 2 were tried, never 2 itself.
f(x) = x + 1 at inputs below 2 and above 2. Below, the outputs climb toward 3; above, they fall toward 3. No column is x = 2.
A function with a hole
Now take . At x = 2 the top and the bottom are both 0, and is not a number, so f(2) is undefined.
Near 2 it is perfectly well behaved: f(1.9) = 3.9, f(1.99) = 3.99, f(1.999) = 3.999, and f(2.1) = 4.1, f(2.01) = 4.01, f(2.001) = 4.001. The outputs approach 4 from both sides, so the limit of as is 4.
The reason: , so for every x except 2 the factor x − 2 cancels and f(x) = x + 2. The graph is the line y = x + 2 with one point missing, at (2, 4). The limit asks only about points near 2, so the missing point does not stop it.
The values of close in on 4 from below and from above. At x = 2 itself there is no value.
The graph of : the line y = x + 2 with a hole at (2, 4). The line runs into the hole from both sides.
The value at the point plays no part
Define g(x) = x + 2 for , and g(2) = 7. Near 2, g agrees with x + 2, so its outputs still approach 4: the limit of g(x) as is 4, even though g(2) = 7.
So a limit can exist where the function has no value, and where it has a different value. The limit says where the outputs are heading; the value says where the function actually is at that one input.
g(x) = x + 2 everywhere except x = 2, where g(2) = 7. The line still runs into the hole at (2, 4), so the limit at 2 is 4; the filled point at (2, 7) is the value.
When there is no limit
A limit needs one number for the outputs to approach. has none at x = 0: h(0.1) = 100, h(0.01) = 10 000 and h(0.001) = 1 000 000, and the negative inputs −0.1 and −0.01 give the same values. The outputs grow without bound, past every number you name, so has no limit as .
A function can also fail by heading to two different numbers from the two sides. That is the subject of One Sided Limits.
Limits of sequences
The same idea works for a list of numbers as n grows without bound. is 0.1 at n = 10, 0.01 at n = 100 and 0.001 at n = 1000: it shrinks toward 0, so its limit is 0.
approaches 2, staying just above it. splits as , which is 3.1, 3.01, 3.001 at n = 10, 100, 1000, and approaches 3.
The running totals of are , , , , …: each total is 1 minus the latest term, so the gap to 1 halves every time. The sum approaches 1, and continuing forever, it is 1.
The usual mistakes
Saying there is no limit because the function is undefined at the point. has no value at 2, and its limit there is 4.
Taking the value at the point as the limit. g(2) = 7, but the outputs near 2 approach 4.
Reading the numerator as the limit of . The fraction keeps shrinking, , , , so its limit is 0, not 1.
Adding the pieces of . The 1 is divided by n, so it shrinks away and the limit is 3, not 4.
Giving 2 for . The totals climb toward 1 and never pass it; a sum reaching 2 would have to start with a term of 1.