FRUIT-GROWING 



I79I 



the following year. This method avoids errors due to difference in the size 

 of the trees, but it has the disadvantage of being based on the judgment of 

 the observ^er, and not on actual weights. This latter method was adopted 

 in the case of one series of observ-ations on trees at Harpenden, and the 

 former in the case of the other series on trees at Ridgmont. 



Whichever method of observation was adopted the results were treated 

 as follows ; two selected trees were numbered i or 2, according to which 

 of them bore the heavier crop in consecutive seasons ; then, comparing the 

 results in two consecutive seasons, if the order of fruiting had been the 

 same (consecutive fruiting) the difference between the numbers given to 

 the trees would be o, if the order had been reversed (alternate fruiting) 

 the difference would be i ; whereas if neither consecutive nor alternate 

 fruiting prevailed, and the results were dependent on chance or on external 

 conditions, the differences wovdd, on the average, be 0.5. An average diffe- 

 rence, for instance, of 0.75 would be halfway between i and 0.5, indicating 

 that the results were dependent on the alternating tendency to the extent 

 of 50 per cent., and on chance fruiting to the extent of 50 per cent. Some 

 of the results obtained by tliis method are given in Table I. 



Table I. 



There can be no doubt but that the results at Harpenden and Ridg- 

 mont are very dift'erent, -but the Harpenden results must be discounted 

 to a certain extent as the number of instances available is smaller. 



In Table II the yields from the Stirling Castle, Bramley and the Va- 

 riety plantation are given, the numbers giving the relative magnitude of 

 the crops compared, in the case of Stirling Castle, with the crop in 1900 as 

 100, and, in the other cases, with that of 191 1 as 100. 



A plus or minus sign has been placed after the values showing whether 

 they are above or below the means of the preceding and succeeding seasons, 

 and the extent to which these signs alternate, though not without some 

 irregularities, is very remarkable. In the case of Cox and Potts no such 

 alternations were recognisable. The cliief factor affecting the alternation 

 of bearing in early flowering varieties is the incidence of spring frosts. 



