Steps / Historical AI

Leonhard Euler · 1770 (1828 English ed.) · Internet Archive

Elements of Algebra

All 627 leaves · leaf 582

Transcription

CHAP. V. ADDITIONS. 545 on account of the indeterminate number $m$, that this expression of $x$ must comprehend all the values that can be given to $x$, in order to make the proposed formula a square; for whatever be the square number, to which this formula may be equal, it is evident, that the root of this number may always be represented by $g + m(x - f)$, giving to $m$ a suitable value. So that when we have found, by the method above explained, a single satisfactory value of $x$, we have only to take it for $f$, and the root of the square which results for $g$; and, by the preceding formula, we shall have all the other possible values of $x$. In the preceding example, we found $y = \frac{5}{3}$, and $x = -\frac{2}{3}$; so that, making $g = \frac{5}{3}$, and $f = -\frac{2}{3}$, we shall have $x = \frac{19-10m-2m^2}{3(m^2-13)}$ which is a general expression for the rational values of $x$, by which the quantity $7 + 15x + 13x^2$ may be made a square. 58. *Example* 2. Let it also be proposed to find a rational value of $y$, so that $23y^2 - 5$ may be a square. As $23$ and $5$ are not divisible by any square number, we shall have no reduction to make. So that making $y = \frac{p}{q}$, the formula $23p^2 - 5q^2$ must become a square, $z^2$; so that we shall have the equation $23p^2 = z^2 + 5q^2$. We shall therefore make $z = nq - 23q'$, and we must take for $n$ an integer number, not $7 \frac{2}{3}$, such, that $n^2 + 5$ may be divisible by $23$. I find $n = 8$, which gives $n^2 + 5 = 23 \times 3$, and this value of $n$ is the only one that has the requisite conditions. Substituting, therefore, $8q - 23q'$, in the room of $z$, and dividing the whole equation by $23$, we shall have $p^2 = 3q^2 - 2 \times 8qq' + 23q'^2$, in which we see that the coefficient $3$ is already less than the value of $B$, which is $5$, abstracting from the sign. Art. 52. Thus, we shall multiply the whole equation by $3$, and shall have $3p^2 = (3q - 8q')^2 + 5q'^2$; so that making $\frac{q'}{p} = y'$, the formula $-5y'^2 + 3$ must be a square, the coefficients $5$ and $3$ admitting of no reduction. Therefore, let $y' = \frac{r}{s}$ ($r$ and $s$ being supposed prime to NN
confidence 0.99 48 expressions source scan at the Internet Archive

This text was produced by machine and has not been corrected by hand. Where it matters, read the scan. Confidence is the model's own, not an external check.