356 /. W. Mallet on the Atomic Weight of Lithium. 



cautiously added to the fluid in the flask from a pipette furnished 

 with a small glass stop-cock and graduated to the one-fifth of a 

 cub. centim. ; the flask being shaken after each addition of the 

 test fluid until the chlorid of silver had completely separated. 

 424 c. c. of this dilute solution of nitrate of silver were needed 

 to complete the precipitation of the chlorine, = '0424 grm. of 

 silver. 



Hence altogether 10*1278+*0424=101702 grm. of silver had 

 been used. 



101702 : 3-9942 : : 1349'66 (equiv, of Ag) : x 

 x =530-06 (equiv. of LiCl). 



530-06-443-28 (equiv. of Cl)=86-78=equiv. of Li. 



This number agrees sufficiently nearly with those derived 

 from the two former experiments to show that all three are de- 

 serving of confidence. If we take the mean of the three, we 

 shall have the number 86*89 for the equivalent of lithium ; and 

 this may, I believe, be fairly trusted as a closer approximation 

 to the truth than any of the numbers hitherto received, if we 

 take into account the greater scale upon which the analyses have 

 been made, and the difference in the methods pursued. For it 

 will be observed that the effect of the difficulty in determining 

 sulphate of baryta already mentioned (namely the adherence of 

 a little of the salt used for precipitation so as to scarcely permit its 

 removal by washing) will necessarily be to increase the appa- 

 rent per-centage of sulphuric acid in the sulphate of lithia an- 

 alyzed, and hence to give a lower equivalent for the alkali than 

 the true one. But this is the method by which the results hith- 

 erto most relied upon have been obtained. 



The number 86 - 89 on the oxygen scale corresponds to 6*95 

 upon the hydrogen- — thus making the equivalent of lithium 

 almost exactly an even multiple of that of hydrogen, in accord- 

 ance with the analogy which seems to extend further and further 

 through the list of elements, as our knowledge of their atomic 

 weights becomes more exact. 



And further, if we take the mean of the equivalents of potas- 

 sium and lithium, using 86*89 for the latter, we get — 

 488*86 (Marignac) 

 86-89 



2)575-75 



287*87 — almost exactly the equivalent 

 of sodium (287*44) as determined by Pelouze. 



