BOOK XV. III. 9 12 



increase until the rising of the Bear-ward, that is till 

 September 16 ; afterwards the increase is in the size 

 of the stones and the flesh. At this stage if rain 

 follows in actually larn^e quantities, the oil is spoiled 

 and turns into lees. The colour of these lces makes 

 the ohve-oil turn black, and consequently when there 

 is only a tinge of black beginning it contains very 

 little lees, and before any blackness shows none at 

 all. People are quite mistaken in supposing what is 

 rcally the near approach of decay to be the beginning 

 of ripening, and it is also a mistake to imagine that 

 the amount of oil is increased by the growth of the 

 flesh of the oHve, since all the juice is then going jnto 

 a sohd form and the woody interior is getting bigger. 

 It is on this account that oHve-trees are watered 

 most plentifully at this period, but watering, whether 

 done intentionally or occurring from repeated falls 

 of rain, uses up the oil, unless fine weather foUows to 

 diminish the solid part of the berry. For, as Theo- 

 phrastus holds, the cause of oil as of other things is 

 entirely warmth, and this is why steps are taken to 

 produce warmth even in the presses and the cellars 

 by h*ghting large fires. A third mistake is in over- 

 economy, as owing to the cost of picking people wait 

 for the oHves to faH. Those who compromise on a 

 middle course in this matter knock the fruit down 

 wilh poles, so injuring the trees and causing loss in 

 the foHowing year ; in fact there was a very old regu- 

 lation for the oHve harvest : ' Neither strip nor beat 

 an oHve-tree.' Those who proceed most carefuHy use 

 a reed and strike the branches with a Hght sideway 

 blow ; but even this method causes the tree to pro- 

 duce fruit only every other year, as the buds get 

 knocked ofF, and this is no less the case if people 



295 



