Dixon and Marshall — Examination of the Wood of Trees. 361 



It follows from these figures that the cross-section of the streaming 

 protoplasm effective in raising the sap is not more than 3'5 per cent, of the 

 whole cross-section, and about -pg- of the total cross-section of the tracheids. 

 Across tliis small section we may expect a flow approximating to 0'03 cm. per 

 minute (taking Janse's average figure for the streaming in water-conveying 

 cells), and this might be considered as being capable of producing a flow of 

 about -[-'-g- that velocity in the traclieidal channels, or about 0'00165 cm. per 

 minute. 



It has been found (1) that water in coniferous wood, when flowing under 

 the action of its own weight, percolates downwards at the rate of from 

 7-10 cm. per hour, or 0'117-0167 cm. per minute. It is evident that 

 streaming of the protoplasm producing a velocity in the tracheids of barely 

 O'OOlGo cm. per minute could not overcome this leak. Even if we take the 

 higliest velocity of streaming which has been observed in any cell, viz., 

 O'Scm. per minute (instead of that observed by Janse, viz., 0"03cm.), and accept 

 tlie hypothetical enzymic pumping-action, we cannot suppose that the rate of 

 lifting water would nearly balance the leak backwards, much less provide for 

 the raising of the transpiration stream in addition. 



Measurements of the Crass-section of tlie Elements of tlie Wood and of the 

 Percolation of Water in the Wood of Dicotyledons. 



For the sake of testing the applicability of the theory to Dicotyledonous 

 trees, we have made some further measurements. In these cases it is not 

 always easy to distinguish on cross-sections between wood parencliyma and 

 tracheids, so some small error in the estimate of these elements may be 

 anticipated. By making several observations this error is probably eliminated, 

 and can never be large. 



In the following measurements microphotographs were employed, and 

 were cut out and weighed in the same way as the camera liicida drawing 

 already described. 



Where the medullary rays are large the ratios of cells to the tracheae are 

 very variable in different photographs, as a field, giving sufficient magnification, 

 may be largely or entirely composed of one or other sort of element. This 

 uncertainty has been minimized by multiplying pliotos of fields taken at 

 random, and so obtaining a more or less characteristic average, and also by 

 determining the ratio of the area of the medullary rays to the rest of the 

 wood on special photos of small magnification. The greatest divergences 

 from the mean were found in Acer pseudoplafaiins. The results of the 

 measurements are given in Table I. The figures indicate percentages. 



