62 Evistence of Radicals in the Amphide Salts disproved. 
can be founded on community of properties? and, by the advo- 
eates of the new doctrine, composition has been made the object 
of definition, instead of being the basis; thus, agreeably to them, 
a compound is not a salt, because it is made of certain elements ; 
but, on the contrary, an element, whether simple or compound, 
belongs to the class of salt radicals, because it produces a salt. 
Since sulphur, with four atoms of oxygen, SO‘, produces a salt 
with a metal, it must be deemed a salt radical. 
28. In proof that the double chlorides are not united in a way 
to justify the opinion adopted by Bonsdorff, Thomson, myself, 
and others, it is alleged by Graham, “that in such compounds 
the characters of the constituent salts are very little affected by 
their state of union.” 
29. This allegation being, in the next page, admitted to be 
inapplicable in the case of the double cyanides ; an effort is made 
to get over this obstacle, by suggesting the existence of another 
compound radical. But the allegation of the author is erroneous 
as respects various double haloid salts, especially the fluosilicates, 
the fluoborates, fluozirconiates, the chloroplatinates, chloroiridi- 
ates, chloroosmiates, chloropalladiates, &c., all of them compounds 
in which the constituent fluorides and chlorides exist in a state 
of energetic combination, by which they are materially altered 
as to their state of existence. 
30. Evidently the word salt has been so used, or rather so 
abused, that it is impossible to define it, either by a resort to 
properties or composition; and I conceive, therefore, that to 
make it a ground of abandoning terms which are susceptible of 
definition, and which have long been tacitly used by chemists 
in general, in obedience to such definition, would be a “ retro- 
grade movement in the science.” I hope Dr. Kane will pardon 
me for employing the language to which he has resorted, in 
speaking of the opinions of Bonsdorff. 
31. If this doctrine, as it has been stated, is to prevail, I do 
not perceive how it is to be prevented from claiming an incon- 
venient extension. ‘The hydrates, as well as the sulphates, must 
have pretensions to contain salt radicals. Hence in the hydrated 
alkalies and alkaline earths, there would be a compound radical, 
consisting of hydrogen, with two atoms of oxygen, hydroxion, 
and these compounds would be hydroxionides; nor can I con- 
