[ 37 ] 
obferved, proceed according to the method defcribed 
above, till each pair have been touched two or three 
times over. But as this vertical way of touching a 
bar will not give it quite fo much of the magnetic 
virtue as it will receive, let each pair be now touched 
once or twice over, in their parallel pofition between 
the irons (Fig. 5.) with two of the bars held hori* 
zontally, or nearly fo, by drawing at the fame time 
the north of one from the middle over the foutli 
end, and the fouth of the other from the middle 
over the north end of a parallel bar ; then bringing 
them to the middle again without touching the pa- 
rallel bar, give three or four of thefe horizontal ftrokes 
to each fide. 1 he horizontal touch, after the verti- 
cal, will make the bars as ftrong as they can poflibly 
be made: as appears by their not receiving any ad- 
ditional ftrength, when the vertical touch is given by 
a greater number of bars, and the horizontal by thofe 
of a fuperior magnetic power. This whole procefs 
may be gone thro’ in about half an hour, and each 
of the larger bars, if well-hardened *, may be made 
to lift twenty-eight troy ounces, and fometimes more. 
And when thefe' bars are thus impregnated, they will 
give to an hard bar of the fame f ze, its full virtue 
in 
! 
* The fmith’s manner of hardening fteel, whom I have chiefly 
employed, and whofe bars have conftantly proved better than any I 
could meet withbefide, is as follows: having cut a fufficient quantity 
of the leather of oil (hoes into very fmall pieces, he provides an 
iron pan, a little exceeding the length of a bar, wide enough to 
lay two fide by fide without touching each other or the pan, and 
at Ieaft an inch deep. This pan he nearly half-fills with the bits of 
leather, upon which he lays the two bars, having fattened to the 
end 
