IX .
" On the Theory of Parallels .
" By Lieut.-General T. PERRONET THOMPSON , F.R.S. Received August 4 , 1862 .
More than thirty published efforts , from Ptolemy downwards , attest the satisfaction with which the Theory of Parallels would be seen established without merging the difficulty into an axiom .
As many of these are certainly not elementary , it amounts to an admission that any knowledge on the subject would be acceptable , even though it let the necessity of beginning from the axiom with freshmen .
Believing that the generation of the straight line with the impossibility of two enclosing a space , and of the plane with the straight line joining any two points lying wholly in the surface , may be rigidly demonstrated from the property of the sphere , which Plato calls its " perfection , " or the faculty of turning about its centre without change of place , -I am induced to submit whether some light may not be offered by the following : Place two equal circles in the same plane , and let a straight line rest upon them ( spheres and a superincumbent plane might be taken , but it is conceived the other is easier ) .
The centres remaining at rest , let the circles dilate as by inflation , preserving always the equality of the diameters ' to one another .
It would appear to be deducible from the Platonic property , that the motion of any point in the circumference during the inflation must be perpendicular to the circumference , and consequently at any instant to the straight line which touches the circumference .
Also the touching point in that straight line is at any instant impelled in a direction perpendicular to the circumference and to the touching line ; out of which it seems impossible that the points of contact in the circumference and in the touching line should ever separate ; for that would imply a motion other than perpendicular in one or both .
If this was supposed allowed , it would follow from making the 268 circles grow less till they vanish , that the distance of the points of contact in the incumbent line was always equal to the distance of the centres .
We should therefore be presented with rectilinear quadrilateral plane figures , of which , from the equality of opposite sides , and the angles at the two extremities being right angles , all the angles must be right angles .
From which it would be an easy step to the proof that the angles of every rectilinear triangle are together equal to two right angles ; and so the Theory of Parallels be entered by another door , and the twelfth axiom be a deduction instead of a groundwork .
It would be interesting if the Theory of Parallels should be proved traceable to Plato 's property of the sphere .

