Steps / Historical AI

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

Elements of Algebra

All 627 leaves · leaf 594

Transcription

CHAP. VII. ADDITIONS. 557 Dρ of the series $D'$, $D''$, $D'''$, \ldots any more than the half of the following term $D_{\rho+1}$. Then we may make $D_\rho = L$, $n_\rho = N$, $D_{\rho+1} = M$, and $\theta_\rho = \psi$, $\theta_{\rho+1} = \xi$, or $D_\rho = M$, $D_{\rho+1} = L$, and $\theta_\rho = \xi$, $\theta_{\rho+1} = \psi$. We must always suppose, as we proceed, that we have taken, for $M$, the less of the two num- bers $D_\rho$, $D_{\rho+1}$. 71. The equation, $cy^2 - 2nyz + Dz^2 = 1$, will therefore be reduced to this, $L\xi^2 - 2N\xi\psi + M\psi^2 = 1$, where $N^2 - LM = A$, and where $2N$ is neither $> L$, nor $> M$, (abstracting from the signs). Now, $M$ being the less of the two coefficients $L$ and $M$, let us multiply the whole of the equation by the coefficient $M$; and making $v = M\psi - N\xi$, it is evident, that it will be changed into $v^2 - A\xi^2 = M$, in which we must make a distinction between the two cases of $A$ positive, and $A$ negative. 1st. Let $A$ be negative, and $=-a$ ($a$ being a positive number), the equation will then be $v^2 + a\xi^2 = M$. Now, as $N^2 - LM = A$, we shall have $a = LM - N^2$; whence we immediately perceive, that the numbers $L$ and $M$ must have the same signs; otherwise, $2N$ can neither be $> L$, nor $> M$; wherefore $N^2$ will not be $> \frac{LM}{4}$; therefore, $a$ =, or $> \frac{3}{4}LM$; and since $M$ is supposed to be less than $L$, or at least not greater than $L$, we shall have, *à fortiori*, $a$ =, or $> \frac{3}{4}M^2$; whence $M$ =, or $< \sqrt{\frac{4a}{3}}$; and $M < \frac{2}{\sqrt{3}}\sqrt{a}$. Hence, we see that the equation, $v^2 + a\xi^2 = M$, could not exist on the supposition of $v$ and $\xi$ being whole numbers, unless we made $\xi = 0$, and $v^2 = M$, which requires $M$ to be a square number. Let us, therefore, suppose $M = \mu^2$, and we shall have $\xi = 0$, $v = \pm \mu$; wherefore, from the equation, $v = M\psi - N\xi$, we shall have $\mu^2\psi = \pm \mu$, and, consequently, $\psi = \pm \frac{1}{\mu}$; so that $\psi$ cannot be a whole number, as it ought, by the hypothesis, unless $\mu$ be equal to unity, or $= \pm 1$, and, con- sequently, $M = 1$. Hence, therefore, we may infer, that the given equation is
confidence 0.93 69 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.