1)4 Mr. Mi M. Pattison Muir on Clieuiical Classijication. 



atom, and in this pi*oportion only. These elements may there- 

 fore be eni])loyed, in place of hydrogen, in determining equi- 

 valents. We know also of certain compound radicles which 

 combine immediately with hydrogen, or with those elements 

 alluded to in the preceding sentence, only in the proportion of 

 one atom to one atom*; we may therefore employ these com- 

 pound radi^'les in determining equivalents. The amount of an 

 element ^vhich combines immediately with one atom of hydro- 

 gen, or with one atom of chlorine, bromine, iodine, fluorine, 

 methyl, ethyl, &c. is therefore the equivalent of that element. 

 The determination of equivalents evidently supposes that the 

 atomic w^eights of the standard elements are known, also that 

 the number of atoms, in the compound under consideration, of 

 the standard element, and also of that element whose equiva- 

 lent is sought for, can be determined. Before the equivalent 

 of an element can be known, the molecular weights of a num^ 

 ber of compounds and the atomic weights of the elements in 

 these compounds must be determined. Many difficulties have 

 to be encountered in determining molecular and atomic 

 Weights. But until a number of these weights have been de- 

 termined, we cannot satistactorily deduce the equivalent of an 

 element, and consequently we cannot fix the valency of the 

 element. 



13. The molecular weights of those compounds which are 

 ca])able of passing into the gaseous state, at attainable tempe- 

 ratures, without undergoing decomposition, can alone be deter- 

 mined with accuracy in the present state of our knowledge!. 

 HencCj if determinations of molecular weights must precede a 

 knowledge of the valency of an element, it follows that we have as 

 yet no accurate knowledge of the valency of a majority of the 

 elements. Indeed, strictly speaking, we do not know with ab- 

 solute accuracy the valency of any element, because, although 

 we have determined the molecular weights of the haloid com- 

 pounds, or of the hydrides, methides, &c. of many elements, 

 we have not determined (with our present knowledge we Can- 

 not determine) the atomic weights of the constituent elements. 

 We certainly have fixed on numbers representing the maximum 

 atomic weights ; but whether these are the true atomic weights or 

 not we cannot say. In most cases the probability that the ac- 



* Altlioiigh it is scarcely correct to speak of an atom of a compound, 

 vet in the case of certain compound radicles the expression is perhaps ad- 

 inissible : thus CH3 may be regarded as the atom, (CHy)^ as the molecule 

 of methyl. 



t The method for determining the molecular weights of gaseous com- 

 pounds is itself founded, like scientific methods generally, on an hypo- 

 thesis ; this hypothesis has hitherto fultilled all the requisites of a good 

 hypothesis. 



