84 ' TRANS. ST. LOUIS ACAD. SCIENCE. 



box. Its back face was blackened, to prevent a double image. 

 A plate of red glass served to cut off the proper quantity of light. 



The w^atch used in making time observations had been lying 

 idle for two years, and it ran satisfactorily until July 22d, when 

 it suddenly stopped. As it had been previously compared with a 

 chronometer at Hannibal, no observations were lost. It was 

 found that the terrible heat of the previous week* had hardened 

 the oil of its bearings. The watch was roughly cleaned here^ 

 and thoroughly cleaned at Keokuk. Its rate was thereafter con- 

 stantly checked by comparison with good regulators and railroad 

 chronometers. A mistake was made in changing the rate of the 

 watch (which was very great), with a view of making it smaller. 



In making declination determinations, the transit was removed 

 from the declinometer, and was placed on a stand outstide of the 

 tent. In doing this, the whole instrument was probably moved 

 on its stand at Wright City. Arrangements have been made to 

 use a separate instrument as declinometer next year. 



It was found after leaving Memphis that our observations there 

 had been made over the site of a blacksmith's shop. 



HORIZONTAL INTENSITY. 



In these determinations the same instrument (D. 3) was used. 



The instrument was set up in the tent, the tripod standing on 

 wooden stakes driven firmly into the ground, the tripod being 

 first levelled on the stakes by properiy driving the latter, adjust- 

 ing with a hand level. 



The observations made were 



1. A series of oscillations with the intensity magnet. 



2. Two series of deflections, C^- being suspended, and Cg de- 

 flecting, at distances of 1.75 and 2.0 ft. 



3. A series of oscillations with Cj-. 



The magnets were suspended on five silk fibres, which were 

 unchanged during the whole of the determinations. (There were 

 originally six, but one was broken in preliminary work before 

 we set out.) 



The reduction to an infinitely small arc was avoided by mak- 

 ing the arc of vibration small. 



The moment of inertia of the magnet Cg, for various tempera- 



* 104" in the shade on several successive days. 



