110 Dr. E. Turner on some Atomic Weights. 



subject, by reducing oxide of lead to the state of metal by 

 means of hydrogen gas. Taking his two most widely differ- 

 ing results, the equivalent of lead, oxygen being 8, will be 

 103'42 in the one case, and 103*64 in the other. His mode 

 of analysis, though apparently easy and simple, is by no means 

 free from practical difficulty. My experiments v.'ere made by 

 converting the oxide into sulphate of lead, a method, I be- 

 lieve, susceptible, with the requisite precautions, of greater ac- 

 curacy than that employed by Berzelius. After many trials 

 I feel certain that the equivalent of lead is not higher than 

 10.3*6. It is probably somewhat lower; so that 103'5, nearly 

 the mean of Berzelius's experiments, is very near the truth. 

 This point I hope to clear up by renewed experiments, which 

 are rendered necessary by the extreme difficulty of getting 

 oxide of lead in a state of adequate purity. In the mean time 

 103*5 is the nearest approximation which experimentjustifies : 

 it is useless to go beyond the first decimal, because we are ig- 

 norant whether 103*5 is greater or less than the real number. 



The following experiment will test the value of this esti- 

 mate: — If the equivalent of lead be 103*5, then 100 parts of 

 metallic lead should yield 146*38 parts of sulphate of lead. 

 The mean of several closely corresponding experiments by 

 Berzelius is 146*419; and the mean of my own is 146*401. If 

 104 were the equivalent of lead, 100 parts ought to yield 146*16 

 of the sulphate, — a number differing widely from the result of 

 experiment, and much beyond the errors of manipulation. 



Chlorine. — The most satisfactory experiments I have met 

 with respecting the equivalent of chlorine are those of Berze- 

 lius. He obtained from 100 of chlorate of potash 39*15 of 

 ox^rgen, and 60*85 of chloride of potassium ; and found that 

 100 of chloride of potassium correspond to 192*4 of chloride 

 of silver. According to my own experiments, 100 parts of 

 silver give 132*8 of chloride of silver, — an estimate extremely 

 close to that of Berzelius. From these data it follows that 

 the equivalent of chlorine is 35*45. 



To compare with this number the equivalent of chlorine 

 determined in a totally different manner, I prepared some very 

 pure chloride of lead, and separated its chlorine by means of 

 nitrate of silver. From the best experiments I could make, 

 100 of chloride of lead correspond to 103*24 of chloride of 

 silver. Now, even taking as the equivalent of lead the theq- 

 retic number 104, the preceding analysis gives 35*578 as the 

 equivalent of chlorine; and when we take the more correct 

 equivalent of lead 103*5, that of chlorine is 35 45, identical 

 with the number deduced from the experiments of Berzelius. 

 The equivalent of chlorine conmionly used by British che- 

 mists, namely 36, is therefore erroneous. 



