NEGATIVE POTENTIAL GRADIENT. 307 



with winds above forty miles an hour. The numbers in the first line of table 1.56 are there- 

 fore only qualitative, but they clearly show that the higher the wind, i.e., the more the drift 

 the higher the potential. The same result is shown in the second line. The numbers in this 

 line were obtained as follows : Th? number of timss that the mean hourly potential gradient 

 exceeded th? hmits of the electrom3ter was counted for each group of winds. This number 

 was then divided by the total number of observations in that group and multiplied by one 

 hundred, the result b^ing the percentage frequency with which the mean hourly potential 

 gradient exceeded the maximum range of tho electrometer during winds of the specified 

 velocity. 



We see that with winds of to 10 miles an hour the maidmum was exceeded in only 

 •7 per cent, of the cases, while with winds greater than 40 miles an hour the maximum was 

 exceeded in 65 per cent, of the cases. In other words excessively high potentials were seldom 

 recorded during light winds, when there was little, if any, drift, while with high winds which 

 were practically always accompanied by heavy drift the potential was as a rule excessively 

 high. 



Drift occurred in two forms : there was first the drift which appeared to fill the atmos- 

 phere darkening the sky and making it almost impossible to see more than a few yards 

 away ; and there was the low surface drift which only extended a short distance from the 

 ground and above which the air was clear and the sky frequently cloudless. Both kinds of 

 drift usually affected the potential in the same way, for as soon as the slightest drift com- 

 menced the potential became verj'' high and remained so until the drift ceased. 



Negative potential gradient. — ^During the whole year the potential gi-adient was negative 

 only on nine days ; as these are important in the following theoretical discussion each will be 

 described in detail : — • 



(1) April 20, 1911. — Throughout this day there was a high S.E. wind of between 30 



and 40 miles an hour. The sky was thickly clouded, and there was either no 

 drift or a very little low surface drift. From midnight to 9 a.m. the potential 

 was positive and high, then it decreased rapidly, and just after 10 o'clock it 

 became negative. For several hoiu's afterwards the needle oscillated fi-om one 

 to the other side of the zero. The potential was never negative for long and 

 high negative values were not recorded. At this time the Sound was not fi'ozen 

 over and there was some raised fog over the open water. The temperature was 

 about -10°F. 



(2) June 1, 1911. — During the early morning there was some new snow and excessive 



drift ; during this period the potential was positive and high. By midday, although 

 the wind was still between 40 and 50 miles an hour-, the drift had entirely 

 ceased. From 13 hours to 16 hours the potential was frequently negative, but 

 only for short intervals, the mean potential being just on the positive side. Later 

 in the day the potential became high on the positive side, but without drift. 

 ■ The sky was overcast throughout the day. 



(3) Septetnber i, 1911.— There was little or no wind on this day. Until midday the 



sky was practically cloudless and the potential was normal. At about 1 p.m. 

 heavy cbuds came up from the north and by 4 p.m. a thick mist lay over the 

 station. With the approach of the clouds the potential decreased and between 

 2 P.M. and 3 p.m. it crossed the zero and was negative for about fifteen minutes. 

 After this the potential rose again and was fairly normal for the rest of the day. 

 There can be Uttle doubt that the cloud was the cause of the reduction and 



