42 MASS. EXPERIMENT STATION BULLETIN 402 



The following material, besides the formulas, has helped in the frost-warning 

 service: 



Imminence of a change to warmer at the end of a cold spell is indicated by the 

 following: 



1 . A substantial rise in the sum of the wet temperature and dew point at inland 

 stations (Worcester, Carlisle, and Paxton) as compared with that of the day be- 

 fore at the same time of day. This indicator sometimes appears with a north- 

 east wind when there is a "Low" along the Atlantic seaboard. 



2. A shift of the wind to southerly (i. e., southwest, south, or southeast) at 

 inland stations (Carlisle, Worcester, and Paxton), especially if the wind is strong. 



3. Much cloudiness inland. 



4. A definite fall of the barometer from 4: to 8 p. m. 



The imminence of the change is certain only when the last of these indicators 

 appears. Some allowance above the reckoning may then be made safely, espe- 

 cially if there is much cloudiness inland. 



Cox has shown^^ that there is a relation between temperature of the top soil 

 and minimum temperature of the air over its surface. Soil temperature, however, 

 is only one of many influences here and is itself largely determined by seasonal 

 factors and the prevailing weather. The writer has observed soil temperatures 

 considerably but has not found them very helpful in computing minimum bog 

 temperatures. See more about this on pages 46, 49, 55. 



On cold nights in late April, very early May, and October, minimum bog 

 temperatures usually fall considerably lower in Middlesex County than in Ply- 

 mouth and Barnstable counties. In most of May and in June, on clear, still, 

 cool nights, minimum bog temperatures tend to be fairly uniform throughout the 

 cranberry district; but in the fall they usually run distinctly higher on the outer 

 Cape and the Vineyard than in Plymouth County and inland. This seasonal 

 difference is due to the change in relationship between the temperatures of the 

 ocean and the land, the former warming in the spring and cooling in the fall much 

 more slowly. Fewer frosts occur on the outer Cape than inland in both spring 

 and fall because there is usually more wind near the sea. The outer Cape is more 

 often frosty on nights with the evening dew point lower at Harwich or East 

 Wareham than inland. Bog temperatures in Pl^'mouth and Bristol counties and 

 inland tend rather considerably to fall lower than those on the outer Cape in the 

 first night of a frosty period but to get somewhat lower on the outer Cape and 

 Nantucket than elsewhere in the last night. The west-east lay of the cranberry 

 region and the usually greater windlness of the Cape and the ocean around it 

 account for this. Nantucket is surprisingly frosty, probably owing largely to the 

 dry and loose character of most of its soil, this being mostly sand with little loam 

 cover. The mosslike growth of Hudsonia ericoides over the treeless rolling up- 

 lands everywhere around the principal cranberry bogs there, by its interference 

 with the transfer of heat to and from the soil, must also have an effect. 



Occasionally in a calm and very cold night in October, after two or more 

 successive cold nights, bog temperatures In the cranberry district fall considerably 

 below the computed minimum^". This is due partly to sharp lowering of the 

 temperature of the soil under continued cold and perhaps partly to reduced con- 

 duction of heat from the lower soil and of the heat of fusion, caused by air pocketed 

 in the soil by freezing ot the surface. Cox noticed this difference when the surface 

 soil froze in Wisconsin. ^^ Ordinary shelter temperatures at 7 p. m. are always 

 near or below freezing when this occurs. 



S^Cox, op. cit.. pp. 42-45, 56-58, and 119. 

 ^''Examples: October 22, 1930, and October 27, 1936. 

 38Cox, op. cit., pp. 41, 47, 79, and 80. 



