172 XINETEENTIl HKK)RT. 



relationships. Neitlier are tlie fall temperatures, according to this 

 method, responsible for variations in the amount of heat required in the 

 spring to bring out the fruit buds, as is proven by column 5, which gives 

 the summations of temperature from the date the fruit ripens until 

 blossoming and shows a range of sums from 48() to 1250. Neither do 

 temperature summations from blossoming to ripening or from January 1 

 to rijDening present much closer agreement. Stated in percentage the 

 relation of the lowest value to the highest, in each column, is as follows : 

 Column 1, .50; column 2 and 3, each 70; column 4, 75; column 5, 38. 

 On this basis, of course, if the sums were the same each year, which is 

 the ideal condition, then the percentage would be 100. 



It is evident, therefore, that this method of studying the heat require- 

 ments of the peach, at least, is not productive of consistent results, and 

 similar studies of other fruits which I have made, but whicli I shall not 

 take the space to note here, yield similarly unsatisfactory results. Smith" 

 has made temperature summations from Mikesell's records for corn, 

 which show as M'ide variations as those for the peach which have just 

 been discussed. He notes a variation from 1232 to 1919 as the least 

 and greatest temperature summations in the twenty-seven years' records 

 considered, during the period from the time the corn plants appeared 

 above ground until they blossomed, and a variation from 897 to 1607 in 

 the thermal values from the dates of blossoming to ripening. As a 

 result of these studies he comes to the conclusion that, "There is little 

 or no relation between the daily mean temperature and the yield of corn." 



The maximum daily temperatures have been considered also as well 

 as the mean temperatures, and Table II, compiled in the same way as 

 Table I, gives the sums of daily maximum temperatures above 42° F., 

 for the peach, the columns corresponding in the two tables. It will be 

 noted that there is a closer agreement between the values given, in each 

 column, in Table II, than in Table I. Stating the relationship between 

 the extremes again in percentage, as was done with Table I, we have 

 in columns 1 to 5, respectively, 64, 71, 72, 78 and 61, or an average of 

 69 per cent, while the average in Table I was 61 per cent, showing an 

 approach of eiglit per cent nearer the ideal method in the latter system. 

 A reason for this closer relationship obtained by considering the maximum 

 temperature readings, may be found in the fact that two days may 

 have the same mean temperature, one dark and cool throughout, with 

 no plant growth, and the other colder at night, but warmer in the day- 

 time, with bright sunshine and considerable growth. The latter would 

 have a higher maximum temperature and hence be given a more nearly 

 accurate value in case the highest instead of the mean tem])erature was 

 considered. 'I'his closer relati()iislii|) betwctii maxinium tciniicratnre and 



