24 



duration to start migration. The notes received came from places where 

 the southerly winds were felt, principally in northeastern Illinois, with 

 the siugle exception of a batch of records from central Iowa, where a 

 northwest wind prevailed. The weather over the whole of the district 

 was clear, with an average barometric pressure of about 30 inches. 



In western Nebraska and eastern Colorado the barometer, though 

 high, was falling, and during the daytime of May 9 it fell rapidly. At 

 Yankton, during the sixteen hours previous to 11 p. in., May 9, the press- 

 ure fell .34 inch to 29.74 inches. Thus the low area developed right in 

 the Mississippi Valley, and its influence was speedily felt, bringing on 

 southerly winds and sending northward almost the last great wave of 

 the spring migration. The temperature did not vary much from that 

 of the night before, except to rise a little in the immediate vicinity of 

 the low area and to fall at La Crosse and northeastward ; the sky was 

 mostly clear ; the wind variable — the low area not having had time to 

 fully affect the winds. All the records came from places where the 

 winds were southerly — consequently from the cloudy places — so that 

 although the larger part of the Upper Mississippi region was clear, the 

 cloudy records form 73 per cent, of the wiiole number. Another thing 

 was noticeable, namely, that notwithstanding the fact that the sky was 

 cloudy, yet the dew point was many degrees below the temperature, 

 showing that the air was very dry. The average of the difference be- 

 tween the temperature and dew point of the records of May 5 is only 

 four degrees, with a range from 0° to 8°, showing that the air was al- 

 most fully saturated with moisture. The records of May 1L show an 

 average difference of fifteen degrees, with a range from 11° to 33°. 

 Thus it will be seen that the humidity of the atmosphere has little or 

 no effect on migration, and can be left out of future investigations. 



This wave of the night of May 9, like the one of five days before, was 

 very extended. Along the western shores of Lake Michigan, where the 

 weather had been cold and disagreeable for the four days previous, it 

 was most strongly felt. Then little or no movement was noted until 

 the Mississippi was crossed. Here, from Keokuk to Moorhead, the 

 night was marked by great activity, and the movement extended in a 

 southwesterly direction as far as southern Nebraska and Kansas. No 

 notes were sent from southern Dakota, and it is probable that almost 

 no migration took place in that Territory, as certainly none did in Man- 

 itoba. Nor was the movement of special importance south of Keokuk. 

 The influence of the low area had not yet extended south of that point, 

 and it was not until the next night that a full bird wave occurred at 

 Saint Louis. The above is an excellent example of a bird wave and a 

 warm wave both working from the north southward. 



Recapitulating, in the same manner as was done for the March notes, 

 it is found that the temperatures at which migration was made are as 

 follows: At 46° there were 29 records of firsts ; at 52°, 11 records ; 55°, 

 116 records; 59°, 66 records; 63°, 70 records; 67°, 9 records. Thus, in- 



