1817.] On Magnetism. 137 
deep red was the Tyrian dye which stained the purple of the 
ancients. Now we know that no species of the genera of vermes, 
murex, or buccinum, gives a colour at all answering their descrip- 
tion ; nor can the colour given by any limax of the present day be 
imitated by cochineal; on the contrary, the Syrian purple of 
Gibbon and Cassiodorus exactly resembles a colour easily made from 
cochineal ; and if the colours of nature be invariable, the porphyry 
of the Byzantine Palace (probably the common criental porphyry of 
a claret red), and the deep bull’s blood red of Cassiodorus, particu- 
larly if, as Gibbon says, these colours could be imitated by cochi- 
neal, must have been much more similar to the colours dyed by 
Bancroft on silk and woollen by madder than to the shell purple 
extracted from the limax inhabiting the murex or buccinum. 
Articte XI, 
On Magnetism. By Mr. Andrew Horn. 
(To Dr. Thomson.) 
SiR, High Wycombe, Dec. 10, 1816. 
Amone those who have sought to develope the principles upon 
which a piece of iron or steel is converted into a magnet, and the 
laws of its phenomena, Epinus and Coulomb are particularly dis- 
tinguished. The theory of Epinus, founded upon mutual attrac- 
tions and repulsions between the magnetic fluid and particles of 
matter, is considered as defective, and has been superseded, at least 
in the estimation of the French philosophers, by that of Coulomb, 
who supposes two distinct fluids combined in the iron when in its 
natural state; but when the magnetic energy is induced, he con- 
ceives the two fluids to be disengaged from their state of combina- 
tion, and, taking contrary directions, are accumulated at the ex- 
tremities of the bar. Thus they exhibit actions analogous to vitreous 
and resinous electricity ; the particles of each fluid repelling one 
another, while they attract those of the other fluid, 
However, | presume that the two fluids, instead of being accu- 
mulated at the extremities of the magnet, circulate in contrary 
directions. On this presumption I shall endeavour to describe the 
process by which a piece of iron is brought to exhibit the magnetic 
virtue ; and the result shall be the test of the theory. Let us con- 
ceive D to be a longitudinal section of a bar of iron or steel in its 
natural state ; the molecule of which we shall suppose to be equi- 
lateral triangles, and arranged asabedefghikimnoparstn, 
and the spaces between thei to contain nearly equal proportions of 
