248 BERZELIUS ON THE ALLOTROPY 
oxide. Moreover, osmium forms lower oxides than the latter. 
These are reduced by hydrogen without heat, whereby the vo- 
latile oxide can be sublimed unchanged in hydrogen, and is only 
reduced when the mixture of the two gases is passed through a 
tube heated to redness, This is in opposition to what ordinarily 
takes place. The higher oxides are usually more readily reduced 
by hydrogen than the lower; here it is the reverse. Can this 
be satisfactorily explained by assuming that the osmium in the 
volatile oxide exists in a different allotropic condition to that in 
the lower degree of oxidation? I will moreover observe, that 
when osmic acid is reduced, by sulphurous acid for instance, its 
lower oxide is not produced, but a blue soluble one, which is in 
a still lower state of oxidation and quite distinct. 
Iridium, when reduced in the moist way by formic acid, is so- 
luble in nitromuriatic acid, according to Dobereiner, That re- 
duced in a dry way, although equally pulverulent, is insoluble 
in it. As it occurs in the mineral kingdom, combined with one- 
fifth of platinum, it has the same specific gravity as platinum ; 
but when obtained artificially, its specific gravity is seldom so 
great as 16°0. Experiments on the specific heat of bodies have 
led to the result, that their atomic weight multiplied by their 
specific heat gives similar products for all; consequently, if the 
atomic weight is the same, the specific heat must also be the 
same. In these experiments no regard was had to the different 
allotropic states, because they were then unknown ; but with car- 
bon this failed, as each modification had.a different specific heat. 
This exception occurs also in iridium. Platinum and iridium 
have the same atomic weight; they must consequently possess 
also the same specific heat. But platinum, the specific weight 
of which is 22°55, according to Regnault has a specific heat of 
‘03243, which nearly agrees with Dulong and Petit’s experi- 
ments; while iridium reduced artificially, of rather less specific 
weight than 16:0, possessed a specific heat of *03683, thus full 
a seventh more. This iridium differs from that naturally occur- 
ring by characters which indicate two distinct allotropic states ; 
and if we add to this what has just been mentioned, that iridium 
is soluble in nitromuriatic acid, it would indeed appear that iri- 
dium exists in three allotropic states ; moreover, in the study of 
its combinations differences are perceptible in its properties, 
which serve clearly to indicate that iridium exists in them under 
three different allotropic conditions. 
