IN reference to my brief letter on this subject appearing on July 18, I should like to say that while the relative values of Curie's constants for iron and nickel agree with those of the analogous ...