ECOLOGY OF ISLE ROYALE. Ill 



east on the third day. The northwest winds prevailed both days of 

 the third wave, while the fourth started with a northwest wind, and 

 changed to west on the succeeding day. The flftli wave was peculiar 

 in having an east wind and the sixth a northeast one, both of which 

 brought birds of different species, and from a different direction than 

 those with northwest winds. These two waves were also much smaller 

 than the preceding ones. It will thus be seen that the great bulk of 

 migration took place with a northwest wind. 



2. Inftiiviicc of Teiiiixratiirc. Since fall migration ])revails at a 

 time when the temperature is gradually falling, the records for a wave 

 near the first of the movement would be mu<'h higher than those at 

 the last, so this factor can be considered only in a relative way, i. e., 

 we must not conqjare the temperature at the first and last of the 

 season, but simply consider the temperature inmiediately preceding and 

 following a wave. The average temperature for the thirteen days was 

 i7°. This low average was partially due to two days of very low 

 temperature. All the \\aves but one were on a falling temperature, 

 and in this lase the mercury had fallen from the day previous. As 

 a falling or low temperature is the cause of the high barometi'ic pres- 

 sure, which in turn with the passage of the high, causes the north- 

 west winds which are so favoi'able to migration, it will be seen that 

 a falling or low temperature is perhaps the first requisite for the bird 

 wave. The low temperature also influences the food of the migrants, 

 killing off the insects, or driving them to shelter, and in this con- 

 nection nmy prove to be very important. 



3. Inftiiencc, of Barometric Pressure. One of the most striking con- 

 ditions was the high baromietric pressure under which these large waves 

 took place. On ten of these thirteen days the baj'ometer stood at 30 

 inches, or above, the average of these being 30.17. The lowest pressure 

 was 29. 8, tlie average for all being 30.09. None of the waves took 

 place oui a falling barometer, but where there was a change the pres- 

 sure was rising as: first wave 29.9, 30.2, 30.3, 30.2; second wave 29.8, 

 30.1, 30.1; third, 30.1, 30.1; fourth 30.2, 30.4. As before stated the 

 direction of the wind is due to the relation of the areas of low and 

 high pressure to the region under consideration, and it is in this con- 

 nection that it bears upon the problem of migration phenomena. 



■ Cooke in his discussion of lie effects of atmospheric changes on 

 spring migration shows tliat at this season, the large movements took 

 place on low or falling barometers, and stated that it probably would 

 be found that in "fall the opposite conditions existed and migration 

 would occur on the rising or high barometric pressures. This was found 

 true at Isle Royale and probably is time for all fall migration. 



4. Condition of the Shy. It will be noticed that on 9 of the 13 

 days of bird waves the sky was clear, and on the remaining four it 

 was simply clouidy, n^o. waves occurring during rainy weather. In 

 spring the waves usually occur during cloudy nights; in the fall, as 

 \vitnessed here, the opposite is the case, and the bulk of the fall migra- 

 tion can be said to take place on clear nights. 



5. Summary and Conclusion. From the data submitted we see that 

 fall migration as witnessed at Isle RoyaJe occurs, in the majority of 

 ca.ses, with a northwest wind and a falling temperature with its rising 



