TRANSACTIONS OF SECTION A. 909 
castle, and were designed, as stated above, to test the working efficiency of the- 
copper wire as compared with iron wires, They were conducted by Messrs. 
Chapman and Eden. 
Copper shows a very decided superiority over iron, the speeds being as 
follows :— 
Copper Tron 
Simplex working » 414 +. ~~, 345 words per minute. 
Duplex a . 270 : p2os  ,, oe 
It is anticipated that the superiority of copper over iron indicated by these 
experiments will have a beneficial and economical influence on our telegraph 
system, and that its extended use will enable us not only to work better, but to: 
dispense with intermediate repeaters in many cases where, on long lines, they are 
now necessary. 
The most interesting point, however, in connection with these experiments is 
that they apparently prove that the superiority of copper is not simply due to its. 
smaller electrostatic capacity and resistance, but that it is more susceptible to rapid 
changes of electric currents than iron ; for when the resistance and capacity of the 
copper and iron wires were equalised by the insertion of resistance coils and con- 
densers, the speed on the former was not thereby diminished. Possibly the 
magnetic susceptibility of the iron is the cause of this. The magnetisation of the 
iron acts as a kind of drag on the currents. It is well known that telephones 
always work better on copper than on iron wires, doubtless for the same reason. 
These experiments also show the high speed of working that is now attained by 
the Post Office authorities with the Wheatstone automatic apparatus. The follow- 
ing table gives an interesting résumé of the different stages of the progress made, 
and its rate of growth :— 
1877 . . 80 words per minute, { 1881. - 190 words per minute. 
1878; 100 ,, - 1982, . 300". 
1879s ..180—,, A TOSS 7 1 be BOO... 4 
BSA 0170) iy, ré TEAM etic: BHO ote, a 
SATURDAY, SEPTEMBER 12. 
The following Papers were read :— 
1. On Orthoptic Loci. By the Rev. C. Taytor, D.D. 
An orthoptie point with reference to a curve is a point at which it subtends a 
tight angle, that is to say, it is the point of concourse of a pair of orthogonal 
tangents. The orthoptic locus of a curve is the locus of intersection of pairs of 
tangents drawn to it at right angles. It is proposed to determine the order of the 
orthoptic locus of a curve of given class. 
The order of the locus is determined when its complete intersection with any 
straight line, for example the line at infinity, is known. 
The tangents drawn from any point on the line at infinity are parallel, and in 
general cannot be regarded as including any finite angle, But the case is different 
with lines drawn from one of the circular points I or J. 
Any two such lines, IO and 10’, may be regarded as including any angle whatso- 
ever, since all the circles that can be drawn through any two points O and O” pass 
through I, 
The orthoptic locus of an ellipse (or hyperbola) is a circle, as De la Hire showed 
ust two centuries ago (Paris, 1685). This may be demonstrated as follows :— 
rom I (or J) one pair of tangents and one only can be drawn to the ellipse,. 
and they may be regarded as intersecting at right angles, Hence I and J being 
single points on the locus and being its only points at infinity, the locus is a circle. 
