Frosts in New York 513 



of its eastern shore, extends in some places probably halfway across the 

 State of Michigan. 



• EFFECT OF ELEVATION ON FROST FREQUENCY 



Every farmer knows that frost is less likely to occur at moderate 

 ele\'ations or on low hills than in low places, and what has been said 

 regarding the increase in temperature at night from the surface upward 

 might seem to afford a sufficient explanation of this fact. However, 

 there are several causes that operate to give elevated lands a greater 

 immunity from frost than adjacent valleys enjoy. A valley usually is 

 shaded for a longer period, both in the morning and in the evening, than 

 are the uplands, and therefore it goes into the night with a smaller store 

 of heat with which to combat the frost. This, however, does not explain 

 the immunity from frost on hillsides that face away from the sun for a 

 large part of the day, which brings us to the consideration of the principle 

 of air drainage. 



The cooling of the air at night begins at the surface, and when the 

 layer of air in contact with the surface becomes cooler than the air 

 about it, it also becomes heavier and begins to slide down the slopes for 

 the same reason that water runs down hill. Hence, the cold air accu- 

 mulates in the valleys. When it leaves the hilltops its place is taken by 

 warmer air, which in turn is cooled by contact with the cold surface and, 

 becoming heavier, starts on its journey to the valley. Thus, a gentle 

 circulation is maintained throughout the night. Again, the stronger 

 winds usually present on plateaus and hills tend to maintain a more 

 uniform temperature by mixing the air and preventing its colder parts 

 from settling to the surface. 



There are many striking examples of the effect of this draining of cold 

 air away from the hills and its accumulation in the valleys. F. W. Stow, 

 of the Royal Meteorological Society, mentions an instance of this character, 

 which shows how remarkably clear-cut the line between frost and no 

 frost sometimes is : 



At the end of May, 1894, there was a frost in the interior of England. 

 After the frost, the ash trees in the valley bottoms were completely frost- 

 bitten; those that grew somewhat higher up on the slopes were frost- 

 bitten below their crowns, but the crowns were green; those that grew 

 on still higher ground had their lower halves frozen, but the upper halves 

 were green; while the ashes growing on the hills surrounding the valley 

 were not injured in the slightest degree. 



Professor Howard, Secretary of the Missouri State Board of Horti- 

 culture, calls attention to the fact that certain sections in that State 

 nearly always escape injury from spring frosts. He instances the famous 



