92 



TRANS. ST. LOUIS ACAD. SCIENCE. 



Change of Magnetic Mo7nent of Cg. 



The magnet Cg seems to have changed much more during the 

 summer than was to have been expected. This was, without 

 doubt, caused by the interaction of the two magnets upon each 

 other prior to July 7th. 



The value of m appears to have reached a minimum about 

 July 10th. It was more nearly constant during the latter part of 

 the summer, but even^ then changed so abruptly as to make an 

 analytical expression almost absurd. 



These facts are exhibited in the annexed table, which gives a 

 reduction of the observed m.agnetic moment w^ to tthe mean tem- 

 perature of the series, which is done by the formula 



m = m, [i -f (/_S9.5) q\, 



where ni indicates the magnetic moment at a temperature of 

 89°.5 on the successive days. The logarithms of these values are 

 given in the next table. For facilitating computation the follow- 

 ing table was used : 



