174 



dated, by shewing that the quantity of 

 branches may be such, as nearly, if not 

 absolutely, to prevent the stems increasing 

 in circumference. On this head the fol- 

 lowing experiments are submitted. 



have yet failed ; nor is it easy to guess when they will,, 

 as they wear much more like metal than wood ; — 

 perhaps, some part of the effect may be imputed to 

 the JVatering ; be that as it may, the circumstance 

 d-eserves attention ; and the more so, because the tree» 

 was very young and free-grown. 



We know of no way by which bends of tolerable 

 seantlings (knees excepted,) can be produced, with 

 certainty, and little trouble; but that sketched on 

 PI. VIII. — The bend is produced from a side-branch 

 only, and kept in that state by the branches of the 

 centre tree overhanging the stems of the others. The 

 former should of course be trained with a flat bushy 

 head. — If fire trees were placed round one centre, and 

 five yards from it, they would stand about six yards 



apart from each other. This hint is given as rather 



curious than useful. Evelyn, Marshall, Nicol, and 



Loudon, have all, in turn, advocated this business; 

 Nature, however, has always, hitherto, frowned disap- 

 probation ; for in this respect, she is Hke the hunted 

 hare; which way be driven from its haunts, but con- 

 stantly returns, or tries to return, so long as life ren>«'iinsv 



