[ *38 ] 



SECTION XIII. 



Experiments Jhewing that no part of 

 plants improve ordinary air, or 

 yield dephlogiflicated air, but the 

 leaves and the green ftalks * 



Quan- 

 tity of 02, THE former experiments 



the two r\ 



airs d e f- w ith flowers, roots, and fruits, are 



troyed. ' 



already above related. There re- 

 main only the green ftalks of 

 branches, not yet covered with the 

 rough fkin or bark, and the wood 

 itfelf, to be examined. 



I put fome green ftalks of a wil- 

 low-tree, the leaves being ftripped 

 off, in a gallon jar filled with pump* 

 water ; the jar was expofed, invert- 

 ed, as ordinary, upon a wall in a 

 warm fun-fhine during four hours- 

 7 They 



