212 . Bulletin 307 



of Group III were read directly from the slide rule, for the data in this 

 group are so heterogeneous and the totals so large that this method is 

 considered sufficiently accurate. 



Comparative value oj different tables. — The tables of Group I are con- 

 sidered of more value than the others for purposes of comparative study, 

 since the averages for the different years are determined from reports of 

 the same orchards for each year, and both yield and income are based 

 on the same orchards. The relation of yield to income and the variations 

 from year to year may be deduced, therefore, with some degree of ac- 

 curacy. On the other hand, in the tables of Group III for 1907 and 1908, 

 together with the two-year and five-year means into which those figures 

 enter, comparison of yield and income for the same point should be made 

 cautiously, since in many cases the figures for yield and income are not 

 based upon the same orchards. This holds true for 1907 far more often 

 than for 1908. 



The value of the income per acre as a measure in orchard statistics. — The 

 value of interpreting results by means of the average income per acre 

 is seen again and again in the present survey. In many cases, if conclu- 

 sions were drawn merely from the mean yield per acre, they would be mis- 

 leading. There is often an unusually heavy yield of inferior apples. If 

 the type of orchard management were judged in this case by yield only, 

 error w^ould be made. For instance, the yield per acre for orchards in 

 sod 5-10 years as shown in Table 22 is 63.2 barrels, which is consider- 

 ably above the general average, 58.4 barrels. Yet the income for the 

 same orchards is only $104, as compared with the general average $111.51. 

 The yield per acre for orchards tilled 10 years or more is 56 barrels, which 

 is below the general average. The income per acre for the same orchards 

 is $128.68, an amount very much above the general average. If the yield 

 alone were considered in these instances, one would conclude that sod 

 treatment of orchards was more profitable than cultivation. A con- 

 sideration of the money value of the crop, however, shows this idea to 

 be erroneous. 



It is the habit for most varieties of apples to bear a hca\y crop nor- 

 mally every other year. This fluctuation in yield per acre is shown in 

 Table i, in which it is seen that 1904, 1906, and 1908 were years of heavy 

 yield, whereas 1905 and 1907 were years of light yield. The incomes, 

 on the other hand, show much less fluctuation. Of the two means, yield 

 and income, the income would seem the better criterion since it is a steadier 

 value. This follows also from Table 2 and the curves of percentage of 

 frequency and distribution. There is much less difference between the 

 mode and the mean in income per acre than in yield per acre. 



