250 BERZELIUS ON THE ALLOTROPY 
are here evident.: The cause of the alteration cannot be in their 
combination, because their tendency to remain combined in the 
oxidized state, judging from their ordinary behaviour, must ra- 
ther increase their affinity for oxygen; it must depend upon 
some other cause; and we can hardly imagine this to be any 
other than a different allotropic condition, similar to that which 
we have distinguished in silicium, chromium and titanium as f. 
If this conclusion be correct, it follows that this allotropic con- 
dition, even although it cannot be demonstrated by art in our ex- 
periments on the isolated elements, may still be proved to exist 
in their compounds, 
Manganese, as we know, after reduction by charcoal in the 
forge, possesses the property of becoming oxidized without the 
application of heat, both in the air and in water, hydrogen being 
at the same time evolved. But Sefstrom has made the remark- 
able discovery, that when it is reduced by silicic acid, so that a 
compound containing 6°8 per cent. of silicium is obtained, a re- 
gulus is formed, which does not differ in appearance from the 
common one, but is not only deficient in the above-mentioned 
properties, but also may be retained at a red heat without be- 
coming oxidized, and is not acted upon by nitromuriatic acid. 
With this great proportion of manganese, this want of affinity 
for oxygen cannot be attributed to the mechanical cause, that a 
covering of silicic acid is formed, which protects that part lying 
beneath it; for platinum, which contains a large amount of 
Silicium, is readily soluble in nitromuriatic acid. The cause 
depends upon the altered properties of the substance ; and it 
appears that silicium, in the state of Sis, has the power of con- 
verting the manganese into the condition Mng, and this at a tem- 
perature at which the manganese itself retains its original allo- 
tropic condition. 
If however this occurs with manganese in the reguline state, 
it must haye a corresponding protoxide, the radical of which 
is Mng, and the silicic acid, with the radical Sis, must also 
change the protoxide of manganese into this condition, and con- 
sequently produce a totally distinct silicate of manganese. Thus 
a new light is thrown upon the nature of the native red silicate 
of manganese (Mn3 Si*), which is unaffected by the strongest 
acids, even when nitric acid forms the oxidizing agent. 
It must not be overlooked, that what is possible with one 
metal, which constitutes one of the most highly positive known, 
