Iu polyatomic organometallic bodies it is remarkable that, with few exceptions, the positive hydrocarbons hold their position much more tenaciously than the associated negative constituents ; and wo thus frequently find the former accompanying the metal through a vast number of compounds ; hence the group formed by the metal and positive hydrocarbons has come to be regarded as a compound radical : thus cacodyl is conceived to be the radical of the whole series of cacodyl compounds : but however great may be the convenience of this mode of viewing organometallic compounds, and the same mode has notoriously been extended to nearly all organic bodies, it must not be forgotten that it is a purely artificial distinction which has no real existence, either in the case of organometallic bodies, or in that of organic bodies in general. A close examination of the habits of the so-called organometallic radicals clearly shows that their atomic power depends upon their position with regard to the stages of stability and maximum saturation ; thus they aro uniatomic when the number of positive groups is ono less than that required to reach either the maximum saturation of the metal, or a lower stage of stability. Cacodyl and tetmmethylarsonium, for instance, are uniatomic radicals, because they are respectively one atom short of the stage of stability and of maximum saturation :— Occasionally en abnormal compound has made its appearance, such as ethoetibylic acid (Si:1C.1100d, or iodide of trietbylstibine
(Sb(C,I1,),I), but further research has invariably demonstrated the incorrectness of such formula and the conformity of the bodies with the normal inorganic types. Indeed this Law may now be regarded as so firmly established, as to be applicable to the control of the formula; of new orgauometallic bodies. From this point of view it is in teresting to watch the effect of the substitution of cblorous or negative by baaylous or positive radicals in metallic compounds. Such a sub stitution affords striking evidence of the dependence of the chemical character of a compound upon that of its individual constituents; a It is obvious, that under favourable circumstances, a compound radical, the number of whose positive atoms is below that of a stage of stability, can have a double atomic character. Thus cacodyl is some times uniatornie, as in oxide of caeodyl, and sometimes teratomic, as in cacodylic acid (As(C,H,)0,). Again, arsenionionornethyl (As is biatomie in arsenious dioxymethido and quadratomic in manomethyl arsenic acid.