NIPHER MAGNETIC DETERMINATIONS IN MISSOURI, 'So. 339 



ago by observation of Polaris on elongation, and has been re-examined 

 each year in the regular work of instruction. 



In the former report it was mentioned that the brass-work of 

 Declin. No. 3 U. S. C. S. contained iron, and the correction to be 

 applied in intensity determinations was given.* It remained to 

 determine the effect of the brass-work upon the position of the 

 needle. For this purpose the magnet box was clamped in such a 

 position that the brass-work could be put on and removed with- 

 out disturbing the needle. The brass pieces were then put oa 

 separately and the effect of each piece noted. As the positions 

 of the side screws which fasten the doors of the box were not 

 changed during the summer, the correction to be applied is easily 

 obtained. It was also important to determine whether the change 

 of the leveling screws in leveling up the instrument had any effect. 

 To determine this, the plates carrying the screws were first put 

 in place, with the screws down to the lowest limit, and they were 

 then screwed up into position. The results with magnet Cg are 

 given below. 



Simultaneous readings of magnet No. i were made, this mag- 

 net being suspended in the University magnetometer, and Cg was 

 corrected for hourly change. Two series first made gave discor- 

 dant results, which were traced to torsion eftects in a new silk 

 fibre used for suspension of Cq. After the magnet had been sus- 

 pended for 24 hours, the fibre became quite constant. The means 

 for six careful and closely agreeing determinations are given. 



Effect of magnetic brass-work of Declinometer No. 3 on Magnet C^. 

 Scale value 2.„-. 



All brass away 



Leveling screws on, but down 

 Leveling screws up in place . . 



-f- Mirror and support 



4- Side screws, all brass on .. 



Page 134, vol. iv. No. i, of these Transactions. 



