
Consider a partition of the interval
![[0,t]](https://img.qammunity.org/2018/formulas/mathematics/college/1vqvrrwjnwmcvf5l3ywkzq92y8spovu2ns.png)
such that
![\displaystyle[0,t]=[x_0,x_1]\cup[x_1,x_2]\cup\cdots\cup[x_(n-1),x_n]=\bigcup_(k=0)^(n-1)[x_k,x_(k+1)]](https://img.qammunity.org/2018/formulas/mathematics/college/onuopqdubcskgmw27axizv15sf9ie5478c.png)
Within each subinterval, select
![\hat{x_k}\in[x_k,x_(k+1)]](https://img.qammunity.org/2018/formulas/mathematics/college/26cnvphons8m4l012mjybrdzqme9ssglt7.png)
so that the value of
![f(\hat{x_k})=\max\{[x_k,x_(k+1)]\}](https://img.qammunity.org/2018/formulas/mathematics/college/s2c8w5nmephmf34lu59s6hf2l8gccunbm9.png)
. In other words,

is specially chosen to guarantee that

is the maximum value of

within each corresponding subinterval.
So now we have an upper bound for the integral in terms of a Riemann sum,

with

. In the limit, these two will be exactly equal:

The main point here is that the right hand side is the sum of the terms of a convergent sequence

as

.
But

diverges if

, so the integral must also diverge.