4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. »9 



In my first-constructed kampometer, I rolled a bimetallic strip, 

 composed of brass and invar in about equal thicknesses, until the 

 total thickness was reduced to 0.008 millimeter. I cut the strip as 

 well as I could with scissors to a width of about 0.8 millimeter. I 

 wound from it a pair of close spirals of opposite curvature, each of 

 about 0.7 millimeter diameter. In each spiral there were 14 complete 

 turns, with brass on the outside. Before cementing them to the glass 

 rods, as described above, each spiral or curl was painted with lamp- 

 black in alcohol and shellac suspension. The total weight of the 

 suspension, including mirror, glass rods, and bimetallic curls, was 

 approximately 4 milligrams. 



The suspension was mounted, as stated above, in a glass tube. A 

 ground-glass cone joint at the top enabled one to rotate the system 

 with respect to the magnetic field. Opposite the upper bimetallic 

 curl was a window of potassium iodide, a substance highly transparent 

 to rays of great wave length. The glass work was very kindly blown 

 for me by Doctor Smith of Mount Wilson Observatory, and the 

 potassium iodide for the window was kindly given me by Doctor 

 Strong of the California Institute of Technology. I found great 

 difficulty at first in sealing the window onto the glass, because 

 potassium iodide fractures so easily under the strain of slight ine- 

 qualities of temperature. Tight sealing was at length secured by 

 using " Arrowhead Cement," a quick-drying water-impervious cement 

 manufactured by the Webb Products Company in San Bernardino, 

 Calif. 



With my colleague, L. B. Aldrich, I mounted the kampometer in 

 the Smithsonian laboratory on Mount Wilson in direct connection 

 through stop cocks and drying tubes with a mercury-vapor pump. 

 We used it at a pressure of 0.003 millimeter of mercury. 



Owing to the crude construction of the suspended system it was 

 impossible to get it perfectly straight. Thus its moment of inertia 

 was much greater than necessary. Moreover, because one of its 

 metals, invar, was magnetic, the slightly crooked system gave in effect 

 still another pair of suspended magnets besides the two principal 

 groups in the control fields. Accordingly there was more than one 

 position of equilibrium. At highly sensitive adjustments when illumi- 

 nated by too strong a beam of radiation, the system would easily 

 reverse itself and remain so. 



Nevertheless, despite these drawbacks due to the crudeness of the 

 construction, the kampometer proved highly sensitive. On August 11. 

 1932, with a time of single swing only f of one second, a candle 

 at I meter produced a deflection of 116 millimeters on a scale at 1.2 



