On some reactions of the salts of lime and magnesia and on the formation of gypsums and magnesian rocks