05.01.2013 Views

RePoSS #11: The Mathematics of Niels Henrik Abel: Continuation ...

RePoSS #11: The Mathematics of Niels Henrik Abel: Continuation ...

RePoSS #11: The Mathematics of Niels Henrik Abel: Continuation ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

148 Chapter 7. Particular classes <strong>of</strong> solvable equations<br />

algebraically solvable if y0 was considered known. Although the equation <strong>of</strong> degree<br />

m<br />

m−1<br />

∏ (z − yu) = 0 (7.12)<br />

u=0<br />

giving the coefficients <strong>of</strong> (7.11) would generally not be algebraically solvable, ABEL<br />

next proved that equation (7.12) ‘inherited’ the property <strong>of</strong> commutative rational de-<br />

pendence among its roots, which φ (x) = 0 possessed. Thus, a ‘descent’ down a string<br />

<strong>of</strong> equations was made possible.<br />

ABEL’S pro<strong>of</strong> <strong>of</strong> this ‘inheritance’, the commutative rational dependence among<br />

the roots <strong>of</strong> (7.12), ran as follows. <strong>The</strong> hypothesis was that all the roots were given<br />

rationally in a single root, i.e.<br />

xu = θu (x0) for 0 ≤ u ≤ m − 1. (7.13)<br />

<strong>The</strong> expression for yu which in the previous argument was given by (7.3),<br />

�<br />

yu = f xu, θ (xu) , . . . , θ n−1 �<br />

(xu) for 0 ≤ u ≤ m − 1,<br />

under the current hypothesis became<br />

�<br />

y1 = f θ1 (x0) , θ (θ1 (x0)) , . . . , θ n−1 �<br />

(θ1 (x0)) .<br />

Combining this with the hypothesis <strong>of</strong> commutativity <strong>of</strong> the functions θ and θ1, ABEL<br />

found<br />

y1 = f<br />

�<br />

�<br />

θ1 (x0) , θ1 (θ (x0)) , . . . , θ1 θ n−1 ��<br />

(x0) .<br />

<strong>The</strong>refore, y1 was a rational and symmetric function <strong>of</strong> the sequence <strong>of</strong> roots (7.13) and<br />

could therefore be expressed rationally in y0 and known quantities. Obviously, ABEL<br />

could carry out the same argument for any other y2, . . . , ym−1. When he let λ (y0) and<br />

λ1 (y0) denote any two among the quantities y0, . . . , ym−1, he found that, without loss<br />

<strong>of</strong> generality,<br />

y1 = λ (y0) = f<br />

y2 = λ1 (y0) = f<br />

�<br />

θ1 (x0) , θ (θ1 (x0)) , . . . , θ n−1 �<br />

(θ1 (x0)) and<br />

�<br />

θ2 (x0) , θ (θ2 (x0)) , . . . , , θ n−1 �<br />

(θ2 (x0)) .<br />

Inserting θ2 (x) for x0 in λ (y0), which transformed y0 into y2, ABEL obtained 10<br />

λλ1 (y0) = λ (y2) = f<br />

while inserting θ1 (x) for x0 in λ1 (y0) produced<br />

λ1λ (y0) = λ1 (y1) = f<br />

�<br />

θ1θ2 (x0) , θθ1θ2 (x0) , . . . , θ n−1 �<br />

θ1θ2 (x0)<br />

�<br />

θ2θ1 (x0) , θθ2θ1 (x0) , . . . , θ n−1 �<br />

θ2θ1 (x0)<br />

10 Here I deviate from my usual notation by writing the composition <strong>of</strong> functions in multiplicative<br />

mode, i.e. θ1θ2 (x0) instead <strong>of</strong> θ1 (θ2 (x0)).<br />

,<br />

.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!