IV.] VARIOUS THEORIES, &c. 131 



on the radial walls only, this neighbouring tracheide 

 will be in a certain direction only i.e. either above or 

 belozu the level of the medullary ray-cell concerned. 



But, it results from R. Hartig's researches that the 

 air-pressure decreases upwards, whence the momentary 

 increase of pressure in the tracheide just referred to 

 will give the water an elastic jog upwards. 



Now suppose that renewed decompositions 

 molecular splittings, &c., again occur in the medullary 

 ray-cell : its osmotic attraction again increases, to 

 repeat the above phenomena, and so on. If the 

 pressures in the various tracheides in contact with 

 the medullary ray-cell differ at any given moment, 

 of course the movements are different, for the same 

 moment, in each. 



Another point should be noted. When the bulged- 

 out closing membrane of the pit suddenly flattens as 

 the cell loses water, it tends to cause a partial va- 

 cuum in the tracheide concerned ; hence it exerts a 

 suction-valve action, and water rushes in from the 

 tracheide below where the air-pressure is somewhat 

 greater. 



If we remember that each medullary ray consists 

 > of numerous cells, each of which is in contact with 

 several tracheides ; and that the number of medul- 

 lary rays is very large, it is clear that these small 



K 2 



