i6o A TREATISE OF 



it ^ for, as I have before obferved, their 

 natural Ihape is much reverfe to a hori- 

 zontal pofition : and when branches are 

 fo laid, they are as much difpofed to put 

 forth ftrong collaterals from their lower 

 buds, as they are long (hoots from their 

 extremities j and when the (hoots are many 

 in number, thofe at the extreme parts muft 

 certainly be (liorter than if there had been 

 no other but them upon the ftem. 



It appears by Fig. 4. that I leave the 

 items, the firft fpring after planting, of an 

 uncommon length. There are many ad- 

 vantages obtained by this method, more 

 than when they are cut (hort ; for thofc 

 branches proceeding from the former will 

 be at a much greater, and a more proper 

 diftance from each other y and a wall may 

 be covered much fooner thereby. The hfe 

 of a tree alfo is much fafer, by the ftems 

 being of fuch a length, than if they were 

 fliorter ; for thofe buds at the upper end of 

 a branch are always larger, and by the 

 fap's motion ea(ier opened, than the buds 

 at the lower end, which appear lefs, and 

 are more compaft. 



By 



