NliVEMBEK 1. 1895.] 



SCIENCE. 



581 



Even a small amount of moisture, say one- 

 half inch of rainfall, will give ample protec- 

 tion if well distributed and precipitated with- 

 in the 24 hours previous. But when severe 

 drouth conditions are prevalent, injurious 

 frosts may occur when the telegraphic tem- 

 peratures do not show a reading within ten 

 degrees as low as in the first case. 



I believe that when estimating the proba- 

 bility or severity of frost sufficient weight 

 has not been given to the dryness or wet- 

 ness of the soil and the resultant dissipa- 

 tion or conservation of heat, and I call 

 special attention to the point as one of the 

 means for improving the forecast. 



I have in mind two typical cases. In the 

 first a high-pressure area attended by clear 

 and cool weather drifted from the west- 

 ward until it covered the State. No rain had 

 fallen with the passage of the low-pressure 

 area immediately preceding it ; hence the 

 ground was in excellent condition for the 

 rapid loss of heat during the night, and a 

 consequent lowering of the temperature of 

 vegetation to the freezing point. Consid- 

 erable damage was done to cranberries in 

 unflooded marshes. In the second case a 

 high-pressure area of slightly greater weight 

 and slightly lower temperature covered the 

 region about ten days later, but it was pre- 

 ceded within a few hours by a light but 

 well distributed fall of rain, averaging about 

 one-half an inch, and no frost occurred. In 

 both cases the wind was gentle from the 

 northwest, and the nights were clear. With 

 slightly lower air temperature and higher 

 barometer in the second condition, heavier 

 frost would have occurred than in the pre- 

 ceding case had it not been for the thinly 

 spread moisture of the timely rain conserv- 

 ing heat at the surface of the earth. 



Might not this principle be carried further 

 in the improvement of the forecast ? As- 

 suming that the caloric energy of the sun is 

 a constant factor, the earth receives each 

 year the same amount or intensity of heat, 



and as the atmosphere is wai'med mainly 

 by contact with or radiation from the earth, 

 seasonal variations of temperature which 

 are marked departures from the normal 

 might result from abnormal terrestrial sui'- 

 face conditions with respect to the conserva- 

 tion of this constant solar energy over large 

 continental areas. Here the excessive or 

 deficient rainfall during the preceding sea- 

 sons should receive careful consideration. 

 The subject is one that requires deeper and 

 more detailed investigation than the length 

 of this paper will permit. 



I find that the minimum temperatures in 

 cranbeiTy marshes during abnormally dry 

 seasons often fall 15 degrees below the tem- 

 peratures telegraphed from the cities and 

 towns within a few miles of the marshes. 

 This is due to the fact that when the loose, 

 spongy peat, of which the marsh is com- 

 posed to the depth of several feet, has 

 dried out, the radiation of heat during the 

 night is very rapid and i^ not counter- 

 balanced by conduction and connection 

 from the marsh. The temperature, there- 

 fore, in cranberry marshes is at all times 

 much lower than that which obtains in 

 marshes composed of heavy black muck, 

 where it presei'ves a more equable condition, 

 such as is common to air resting over a con- 

 siderable body of water. A dry cranberr}' 

 marsh does not, therefore, enjoy that im- 

 munitj' from frost which wet marshes and 

 watery lands get the benefit of. But when 

 the ditches are flooded from the reserve 

 water supply on receipt of a frost warning, 

 the water quickly percolates through the 

 peat composing the marsh, and the rapid 

 loss of heat by radiation is checked and the 

 frost averted. 



The degree of heat to which vegetation 

 has been subjected immediately before the 

 frost condition, and the temperature under 

 which it had made its growth, will in a 

 great measure determine the extent of 

 damage to ensue. 



