13 .
" On the Analysis of Numerical Equations .
" By J. R. Young , Esq. , Professor of Mathematics in Belfast College .
Communicated by Sir John W. Lubbock , Bart. , F.R.S. &amp;c. The object of this communication is to diminish the labour attendant upon existing methods for the analysis of numerical equations .
As Budan pointed out intervals , within the bounds of the extreme limits of the roots of an equation , in which all search for roots would be fruitless , so here the author seeks for what he terms " rejective intervals " among those which Budan had retained .
This he proposes eff'ecting by transforming the first member of every equation X=0 into X={F-t VP-X } x{FvI-X } ... .
( 1 .
) which the author calls decomposing it into con ; jugate factors ; in which factors , F is entirely arbitrary .
On account of this unrestricted character of F , innumerable changes may be effected on a pair of conjugate factors , without disturbing their product : but it is from the following expression of these factors that the general results in this paper are chiefly deduced : namely , X= { F+f+ F+2Ff+f X } .
( 2 .
) x{ F+f/ 'F + 2Ff+f2 -x J from which it follows that , having decomposed any function X into a pair of conjugate factors ( 1 .
) , we may always afterwards add any quantity , f , to the rational part of each factor ; provided we , at the same time , introduce the expression 2Ff+f2 under the radical : and it is upon this general truth that the results in the paper entirely depend , and from which the rejective intervals are discovered .
The value of the principle is illustrated by examples taken from STURM , FOURIER , and others ; and some general theorems are deduced-applicable to all equations-in reference to the existence of imaginary roots , which furnish some remarkably simple criteria .
For instance : it is shown that if , in the general equation of the fourth degree , 4+ ax + +bx+ cx + d=O in which d is positive , the condition 4(b-a0)d &gt; c2 exists , all the roots are necessarily imaginary .
And that if , in the general equation of the sixth degree , x6 + ax5 + bx4 + CX3 + dx ex + f= , in which f is positive , the condition 4{ d(b --a ) 4c }f &gt; ( ( b-a)e2 exists , each member being positive , all the roots are necessarily imaginary .
The paper concludes with some general propositions , derivable from the principles established in the preceding investigations , and which the author conceives to be of value in the analysis of equations .

