of hot iron between the white and blood-red heat . 1 25 
It should be observed, that all the above experiments 
were made with the bars inclined in the direction of the 
dipping needle, or nearly in that direction, and it will be seen 
that the negative attraction was the greatest where the 
natural attraction was the least ; that is, opposite the middle 
of the bar, or in the plane of no attraction. I was led, there- 
fore, to make a few experiments with the bar inclined at 
right angles to its former position, but the results were by 
no means so strongly marked as in the preceding experi- 
ments : we always found a certain quantity of negative at- 
traction, but it was very inconsiderable, never amounting to 
more than 2i°. 
I also made one experiment with an iron 241b. ball, but 
the heat was too intense to make any very accurate obser- 
vation. The numbers obtained were 
Cold attraction +13 0 30'. White heat o° o'. 
Red heat — 3 0 30'. Blood-red heat -f- 19 0 30'* 
It may be proper also to state, that, being doubtful how 
far the heat itself, independent of the iron, might be the 
cause of the anomalous action above described, I procured 
two copper bolts of rather larger dimensions than my iron 
bars, and had them heated to the greatest degree that metal 
would bear ; but on applying them to the compass, no mo- 
tion whatever could be discovered in the needle. 
The only probable explanation which I can offer by way of 
accounting for these anomalies, is, that the iron cooling faster 
towards its extremities than towards its centre, a part of the 
bar will become magnetic before the other part, and thereby 
cause a different species of attraction ; but I must acknow- 
ledge, that this will not satisfactorily explain all the observed 
