252 The Principles of Fruit-growing 



likely they are to be injured by low temperatures. While 

 tables giving injurious temperatures to fruit when in 

 bud, blossom and so on, have been prepared, it is safe to 

 say that these temperatures are not entirely reliable. 

 This is because conditions are never the same in any frost 

 period. It may be stated that in the practice of orchard- 

 heating the safest plan is to keep the temperature just 

 above the freezing-point, no matter what the variety of 

 fruit. No doubt, this will often be 2° to 3°, or even more, 

 higher than necessary, but the practice is on the safe side. 

 In orchard-heating practice, the temperature should 

 never be allowed to go much below the danger point, as it 

 is usually difficult to bring it back without some chance 

 of injury to the fruit." 



How to predict frost. 



There is no certain way of foretelling frost. Careful 

 observation of conditions in the locality, long experience, 

 and the use of the forecasts of the Weather Bureau are 

 the most reliable means. Formerly the evening dew- 

 point was considered a reliable guide to the minimum 

 temperature of the ensuing night. Hence, if the dew- 

 point say at 4 p.m., as determined by the wet-and-dry- 

 bulb thermometer, or psychrometer (Fig. 93), should lie 

 below the freezing-point, 32° F., frost was considered 

 likely to occur. This method of forecasting frost seems 

 to hold good for certain localities in the western states, 

 but has been shown by Cox and others to be wholly unre- 

 liable in the more humid eastern states. W. M. Wilson 

 concludes (Cornell Bull. No. 316) as follows, after a 

 study of the subject: "For the farmer who is prepared 

 to -make practical use of a frost-warning, the forecasts 

 issued by the Weather Bureau should receive first con- 



