Dixon and Marshall — Examination of the Wood of Trees. 359 



Janse (9 aud 10) has worked out his theory in considerable detail. As 

 was the case of Q-odlewski's theory, Jaiise's applies most easily to Conifers. 

 He assumes that the protoplasm in the cells of the medullary rays during 

 transpiration circulates actively — the circulation taking place chiefly in a 

 vertical plane. The protoplasm which is moving across the lower horizontal 

 face of tlie cell, he supposes, fixes water in its vesicles by means of an enzyme J 

 the water is retained as the protoplasm climbs up one vertical side, but is 

 released from the vesicles by a reversal of the enzymic action, as the stream 

 reaches the upper horizontal wall. The water fixed in the vesicles on the 

 lower side is taken from a lower tracheid, and when it is released from the 

 vesicles on the upper side, it is liberated into a tracheid at a higher level. 

 The one medullary-ray cell is, at least in the Conifers, supposed to serve several 

 traclieids on each side. 



According to this scheme, at any one moment half the protoplasm of 

 each cell, at the most, is raising water, while the other half is returning 

 empty. Therefore we may assume that the whole upward transpiration 

 stream is being passed through, as a maximum value, half the lumen of each 

 of the medullary-ray cells in any cross-section. Hence, if we know the 

 average velocity of the streaming, and measure the cross-section of the cells, 

 we will have a major limit of the amount of water passed upwards in this 

 manner. To obtain the actual velocity of the stream in the wood caused by 

 this supposed action of the cells, it will be necessary also to make allowance 

 for the amount of water which will be continually flowing backwards through 

 the tracheae under the action of gravity. 



So far as we are aware no measurements have been recorded on the 

 velocity of streaming in the medullary-ray cells or in the cells of the wood 

 parenchyma. Janse (9) has observed the streaming, but has not recorded its 

 velocity. The average velocity of streaming in the endodermis cells, to which 

 he assigns a similar function, he estimates at 0'03 cm. per min. at 19°. The 

 most rapid streaming in closed cells hitherto observed seems to be 0'2-0-3 cm. 

 per minute (5). 



All direct intervention of the living elements of the stem in the lifting 

 of the transpiration stream has been negatived by such a large body of 

 experiment (3) that the refutation of any special hypothesis involving this 

 intervention seems almost superfluous ; however, as this consideration of 

 Janse's has led us to make some measurements on the structure of the 

 conducting tracts of some trees which may be of use to other investigators, we 

 have thought it worth putting these measurements on record, and incidentally 

 discussing their bearing on this hypothesis. 



3 M 2 



