Mr Broun points out the defects of the usual methods of ascertaining the corrections for temperature, applicable to magnetic instruments generally, and the vertical force magnetometer in particular. The usual methods depend upon the knowledge of the time of vibration, or upon the statical deflection produced by a neighbouring magnet, under differing circumstances of temperature. Mr Broun shews that both these methods are liable to great objection; and he has succeeded by arranging the ordinary hourly observations of the instruments in groups, in obtaining consistent results for the temperature correction by a process of elimination.