Physical Properties of analogous Organic Compounds. 195 



Where several substitutions occur successively, it is requisite 

 to calculate from each of the compounds examined the den- 

 sity of the other, so that any accidental error of experiment 

 may not be allowed to have too great an influence. To be as 

 precise as possible, I will here take for calculation two of the 

 best-examined compounds, one containing the greatest num- 

 ber of atoms of hydrogen, and the other the greatest number 

 of atoms of chlorine. 



From this table it becomes evident that the law above ad- 

 vanced is generally confirmed by experiftient. However, the 

 calculated densities frequently differ more from those observed 

 than was the case for the other laws. This arises from various 

 circumstances. 



It is very difficult to obtain by substitution a pure com- 

 pound, one which is not rendered impure by the presence of 

 another preceding or following degree of substitution. This 

 uncertainty in the preparation of the compounds, to which 

 Berzelius has drawn particular attention, is the cause that 

 the statement of various observers are often in contradiction. 

 In the 4- 1st volume of the A)male?i der Chemie and Pharmacie, 

 I have treated at length of the uncertainty and contradiction 

 of the observations on such compounds, and have shown that 

 the differences between the observations are not smaller than 

 those between the calculated and the observed magnitudes. 



Another reason why the calculation does not always agree 

 accurately with the experiment, is that the temperatures at 



2 



