Skip to content

Mathematics · Ch 10 — Sequence and Series

Arithmetic Progression

2

Arithmetic Progression

A sequence a1,a2,a3,…a_1, a_2, a_3, \ldots is called an Arithmetic Progression (A.P.) if the difference between every pair of consecutive terms is the same constant, called the common difference, usually denoted dd:

d=a2−a1=a3−a2=⋯=an+1−anfor every n.d = a_2-a_1 = a_3-a_2 = \cdots = a_{n+1}-a_n \quad \text{for every } n.

If the first term is aa (also written a1a_1), the A.P. is written out as

a, a+d, a+2d, a+3d, …a,\ a+d,\ a+2d,\ a+3d,\ \ldots

so that each term is obtained from the previous one by adding dd. If d>0d>0 the A.P. is increasing; if d<0d<0 it is decreasing; if d=0d=0 every term equals aa.

The nnth (general) term. Reading off the pattern above, the first term needs 00 copies of dd added to aa, the second term needs 11 copy, the third needs 22 copies, and in general the nnth term needs (n−1)(n-1) copies:

an=a+(n−1)d.\boxed{a_n = a + (n-1)d}.

This can be proved formally by induction on nn: it holds for n=1n=1 (giving a1=aa_1=a); and if it holds for some nn, then an+1=an+d=[a+(n−1)d]+d=a+nda_{n+1} = a_n + d = [a+(n-1)d]+d = a+nd, which is exactly the formula with nn replaced by n+1n+1 — so it holds for every nn.

Sum of the first nn terms — derivation by "reverse and add". Let

Sn=a+(a+d)+(a+2d)+⋯+[a+(n−1)d].S_n = a + (a+d) + (a+2d) + \cdots + [a+(n-1)d].

Write the same sum with its terms listed in the reverse order:

Sn=[a+(n−1)d]+[a+(n−2)d]+⋯+(a+d)+a.S_n = [a+(n-1)d] + [a+(n-2)d] + \cdots + (a+d) + a.

Now add these two expressions for SnS_n term by term. In each column, one term contributes a+kda+kd from the first line and a+(n−1−k)da+(n-1-k)d from the second, and these add to exactly 2a+(n−1)d2a+(n-1)d — the same value in every one of the nn columns, because the dd-parts always add up to (n−1)d(n-1)d regardless of kk. So

2Sn=n[2a+(n−1)d]⟹Sn=n2[2a+(n−1)d].2S_n = n\big[2a+(n-1)d\big] \quad\Longrightarrow\quad \boxed{S_n = \frac{n}{2}\big[2a+(n-1)d\big]}.

This is the celebrated trick attributed to the young Gauss, who is said to have summed 1+2+⋯+1001+2+\cdots+100 instantly this way.

An equivalent form using the last term. If l=an=a+(n−1)dl = a_n = a+(n-1)d is the last (i.e. nnth) term of the sum, the formula above can be written as

Sn=n2(a+l),S_n = \frac{n}{2}(a+l),

i.e. nn times the average of the first and last terms — since 2a+(n−1)d=a+[a+(n−1)d]=a+l2a+(n-1)d = a + [a+(n-1)d] = a+l. This form is often quicker to use when the last term is already known, and both forms must always agree.

Recovering a term from consecutive sums. As noted in Section 1, an=Sn−Sn−1a_n = S_n - S_{n-1}; for an A.P. this gives an independent check on the sum formula, since substituting Sn=n2[2a+(n−1)d]S_n=\frac n2[2a+(n-1)d] and Sn−1=n−12[2a+(n−2)d]S_{n-1}=\frac{n-1}{2}[2a+(n-2)d] and simplifying does indeed recover an=a+(n−1)da_n=a+(n-1)d. …