220 Prof. J. B. Farmer. On the Quantitative Differences in the 



means of a camera lucida at an exact magnification of 10 diameters. By 

 means of a planimeter the area of the wood thus obtained can easily be 

 determined. Inasmuch as the instrument used by me measures areas in 

 inches a reduction to centimetres has to be made, and the result divided by 

 100 gives the real area in square centimetres. 



It may be objected that a section taken from the middle of the stem will 

 not give absolutely accurate data for estimating the flow per square centi- 

 metre through the twig. . But, having regard to the natural taper of the 

 stem and the fluctuating area of the wood, it is probably as near an 

 approximation as it is possible to reach. Furthermore, as all the specimens 

 are treated in the same way, the error, in so far as it afiects the value of 

 comparative results, will in any case be very small, and well within the 

 limit of unavoidable experimental error. From the data obtained as above, 

 the volume transmitted by a stem 15 cm. in length per 1 sq. cm. of wood 

 (in cross-section) is calculated, and this ratio constitutes what is hereafter 

 termed the " specific volume," and it expresses quantitatively the " specific 

 conductivity " of the wood of any given twig. Comparison can be made with 

 other branches, whether of the same or another species, thus enabling a very 

 precise estimate to be made of the range of variation occurring in a par- 

 ticular species, and this again serves as the basis for comparing the 

 behaviour of different species with one another. 



The apparatus employed in this research is shown in the accompanying illustration 

 (fig. 1), for which I am indebted to Miss Reeks, Technical Artist. It consists, 

 essentially, of a tube which conveys the water, under any desired pressure, to the 

 twigs figured as dipping into the four measuring glasses. A special arrangement also 

 renders it possible to suck water up through the twigs from the measuring glasses, and 

 to determine the amount so transmitted by means of the burette marked P in the figure. 



The water is supplied fiom the main supply by means of the tap A ; it passes 

 through a glass wool filter Dj to a T- piece, and thence in the direction of the arrow 

 through the tap E, and a second glass wool filter D2 to the three-way tap F, which 

 connects with the manometer. The water rises in the tube leading to the latter, but is 

 prevented from passing into the manometer head by the two pairs of bulbs Gj and Gj. 

 The three-way tap F further carries the water through H to the four twigs through the 

 three-way taps Jj, Jg, J3, J4. Any or all of these twigs can, of course, be disconnected at 

 pleasure from the water in the horizontal tube leading to the three-way tap K, which 

 ordinarily serves as a cut-ofi^, and thus all the water is delivered through the twigs, and 

 the amount from any or all of them is severally collected in the measuring glasses. Any 

 air bubbles that may occur in the system are easily washed out by opening the taps K 

 and 0, and drawing the water through the apparatus by means of the filter pump B 

 attached to the middle of the three main taps. For this purpose the bottle of water N 

 is specially useful, as a means of completely washing out the whole of the horizontal 

 tube. 



The pressure of water from the main is kept nearly constant by the device of 

 connecting the limb of the T-piece beyond Dj with a filter pump attached to the main 

 tap A. The negative pressure thus produced is roughly adjusted to carry ofi' a part of 



