IV.] VARIOUS THEORIES, &c. ni 



with its air (it cannot retain in solution so much as 

 it could under a higher pressure) and the air will 

 continue to pass off from the sap so long as the 

 pressure is less than an atmosphere. Hence air- 

 bubbles must be produced when more water is 

 passing off at the leaves than is entering at the 

 roots. 



The reply that the water held fast by the air- 

 bubbles proves too much, because it is immovable, is 

 anticipated by the remark that though each column 

 is thus fixed as a whole, the individual particles of 

 water are free to move. 



We have already examined Elfving's proofs that 

 the water travels through the lumina, and may at 

 once pass to his conclusions. He points out that the 

 imbibition theory was devised to meet the difficulty 

 that water is carried up several hundred feet in the 

 tallest trees, whence the water (i) must be held by 

 molecular forces, (2) must be easily moved : but the 

 theory of intra-cellular water above stated quite agrees 

 with this. Transpiration at the leaves is supplied by 

 osmotic currents from the extremities of the vascular 

 bundles to the mesophyll : these osmotic currents are 

 feeble, but they set the water in movement, via the 

 vascular bundles of the ribs and petiole. As soon as 

 the stream encounters an air-bubble, it bends to one 



