190 Account of the Levelling taken from 
immersed, and which easily dissolves the oxide of iron formed 
by the decomposition of the silex, thereby facilitating the reduc- 
tion. Whereas, when the iron is fused by itself, the silex can 
only be furnished by the crucible, to which it coheres with con- 
siderable force; and the oxide of iron, as it forms, soaks into 
the crucible, and serves to protect the earth from the contact of 
the metal ; which is doubtless the cause why the conversion into 
steel cannot be completed without the presence of a glass. 
We cannot, therefore, judge of the degree of heat required for 
the fusion of iron in a Hessian crucible, since it appears demon- 
strated that at a very high heat iron reduces silex, and combines 
with the silicium thus produced into a compound more fusible 
than iron perse. On the other hand, when platina is heated 
with silicium already formed, it unites with it into a more fusible 
compound; but if this metal does not melt by itself in a Hessian 
crucible, it is because it has so littie affinity for oxygen, that it 
has not, like iron, the property of decomposing silex. 
Though we cannot fix the degree of fusion of pure iron, any more 
than that of platina or of manganese, we may at least determine 
their relative fusibilities when in contact with charcoal or silicium, 
or both together ; which, in a crucible lined with charcoal, is in 
the following order; namely, iron, platina, and manganese: and 
if we admit it to be probable that this is the real order of fusibi- 
lity when they are pure, it will follow that manganese is a more 
refractory metal than platina. 
XLIV. Account of the Levelling taken from the Trigonometrical 
Station on Rumbles Moor and the Observatory, to the Canal, 
_,and. ultimately to the Irish Sea; being a Continuation of the 
Article given in our last Number, p. 130. By A Corre- 
_ SPONDENT. 
To: Dr. Tilloch. 
Sir, — Tue usual method of measuring the fall of a declivity is 
by means of a telescopic level placed between two staves marked 
with feet and inches, with a little additional apparatus to en- 
able the observer to raise or depress the cross wires to the nearest 
inch on the first erected staff, and also to alter the height of the 
one in advance until a particular inch is covered by the telescope 
(by which means the fractional parts of inches and the use of the 
sliding vanes may be avoided) :—a more accurate method could 
not be devised. It must, however, be found extremely tedious in 
practice, and the more so in proportion to the abruptuess of the 
descent. Wishing if possible to avoid any errors of an optical 
natule, 
