224. ANNUAL REPORT SMITHSONIAN INSTITUTION, 1926 
this circumstance did not influence the efficiency of the instrument. 
The records gave however, only relative values of the gradient. In 
order to reduce them to absolute values, simultaneous eye observa- 
tions were carried out from time to time on smooth ice at a sufficient 
distance from the ship. As a matter of precaution in case a 
polar bear should be too curious, the observer was always armed 
when he had to go some distance from the ship. I may mention 
that the observers were never disturbed. 
We were unable to secure any observations during the summer 
because a satisfactory insulation could not be maintained on account 
of the dampness of the air. Our records are, therefore, limited to 
the cold months, October to April. When referred to universal 
time, the records for this season are in excellent agreement with the 
results obtained on the Carnegie. These are represented by the lower 
curve in Figure 2, while the three upper curves represent our pre- 
liminary results during the three winters. Our observations from 
the Polar Sea thus confirm the important conclusions regarding the 
universal character of the diurnal variation of the potential gradient 
drawn from the observations carried out on the Carnegie during 
cruises over all oceans. 
The greatest value of the gradient occurs at 18" Greenwich mean 
time, which is approximately the time when the sun is in the meridian 
of the magnetic poles of the earth. This fact indicates a close rela- 
tionship between the magnetic and electric fields of the earth, but the 
character of this relationship has yet to be explained. 
Meteorological observations were taken regularly six times daily 
during the three years, and for the entire period continuous records 
of the barometric pressure, the temperature and humidity of the 
air, the direction and velocity of the wind, and the duration of sun- 
shine are available. Our meteorological screen was placed on the 
roof covering the deck, while a snow gauge for measuring the amount 
of precipitation was placed on the ice. Special studies of the humid- 
ity of the air at low temperatures and of the formation of frost were 
carried out by the assistant scientist, Mr. Malmgren, who devised and, 
assisted by Mr. Dahl, constructed a special instrument for recording 
the frost formation. Special studies of the daily variation of the 
temperature of the air were also carried out, but I can not enter upon 
a discussion of the results of these investigations nor of the results 
of the general meteorological observations. Instead I shall turn to 
our upper-air observations. 
The direction and velocity of the wind aloft was determined by 
means of pilot balloons, 552 of which were released. ‘These wind 
observations indirectly give interesting information regarding the 
average temperature distribution at great altitudes. In Figure 3 
average wind velocities in the free atmosphere are represented by 
