on the Atomic Weights of Bodies. 



255 



TABLE I. — Simple Substances. 



•• I have adopted 3-625 for the atom of nickel partly from the experiments of 

 Tuppuli and Rothoff, and partly because Berzelius's number is as high as 3-69675. 

 But I hare never been able, from my own experiments, to obtain so hi°-h an 

 atomic number for nickel. I made a very careful analysis of sulphate of nickel 

 very lately. The salt had been twice crystallized, and had all the characters of a 

 pure and neutral salt. From 100 grains of it I got 



Oxide of nickel, 25-17 4-41 



Sulphuric acid, 28-50 5 



Water, 40-88 



Impurity, 1-75 



96-30 

 The water had it been all extracted would have amounted to a grain more. The 

 loss is almost 3 per cent., and the 1-75 grain of imparity destroys the confidence 

 I expected to DBVI derived from tliis analysis. 



