Royal Society. 131 
Hones cannot fix a pole in a thick fhort key, which yet they 
,will do in a imall key^ fo in a fliort thick iron, tapering, a 
loadltone may fix a pole in the fmall end, when it cannot 
in the great end. \Yhen ignition, hammering, or a loadftone 
cannot make fixed poles, it muft not be thought that it can do 
ablolutely nothing on luch rods 5 for even then it may be 
found, that there is an effect of maguetifm in them difcerniable 
enough otherwiie, tho' not lufficient to make fixed poles. VV hen 
you have the due length for making of a fixed pole, you will find 
that the making one a fixed north, will confequently render the 
other a fixed foath pole 5 but if keeping the fame diameter of 
this rod, you encreaie its length luiSciently the making one end 
a fixed north pole will not neceflarily make the other a fixed 
iouth pole , but leave it a mutable pole ^ fo if you by a like pri- 
mary operation make the fecond end a fixed pole, the firfl will 
loole it's fixednefs and become mutable 5 1 lay, there is a certain 
length fuited to every thicknefs of iron, to leave one end mutable, 
whilfl the other is fixed, and the thicker the iron is, the greater 
is the length ; if you firther encreaie the length of the fame rod, 
you will attain fuch a length, that when you have fixed a pole on 
one end, then go to fix the other end, the fixity uf the firft will 
not be deftroyed, and that end become mutable as before, bu^ the 
fixity of the firft end will remain 3 and io you may make both 
ends two fixed north poles, or two fixed fouth poles ; I fay the 
ihorteil length (there are no limits of the greateft length) for this 
is more in thick, than in thin iron. The aforelaid lengths are 
lefs, according to the ftrength of magnetifm ^ viz. ignition re- 
quires a gteater length than when a rod is actuated by a load- 
Itone, and a rod touched with a ftrong load ftone requires lefs 
length than one touched with a weak one. 
"Toftnd the Sun's Ingrefs into the Tropical Signs ^ by Mr. Edm. 
Halley. Phil. Tranf. N" 2 13. p. 12. 
MR. IMley does here give a method of finding the moments 
of the lolftices, capable of all the exadtncfs the mod accu- 
rate candefire, and that without any confideration of the parallax 
of the fun, of the refra6tions of the air, of the grcateft obliquity 
of the ecliptic, or latitude of the place j all which are required 
to afcertain the times of the equinox from oblervations, and 
which, being faultily affumed, have occafioned an error nearly of 
three hours in the times of the equinoxes, deduced from the ta- 
bles of the noble I'ycho "Brahe and Kepler ^ the vernal being fo 
much liter, and the autumnal fo much earlier than by the cal- 
R 2 cuius 
