SUPPLEMENT 87 



an internode, it may be conveniently cut into longitudinal strips, each of which 

 consists of one kind of tissue only. Epidermis, cortex, wood, and pith are easily 

 separated from each other, and, by measurement, it may be determined what 

 percentage of elongation or shortening each of these tissues undergoes. As- 

 suming the original length to be 100, the following changes in length are shown, 

 e. g. by Nicotiana tabacum (SACHS'S Textbook, p. 297) : 



No. of Internode. Percentage alteration in length. 



I (youngest) epidermis cortex wood pith 



III-IV -2-9 -1-4 +3-5 



V-VI 2-9 1-3 0-8 +2-7 



VII-IX 2-7 2-1 o-o + 3-4 



X-XII 1-4 0-5 o-o +3'4 



XIII-XV -1-05 -o-o ? +4-0 



It will be seen that all the tissues undergo contraction, save only the pith, 

 which elongates. In the uninjured internode, therefore, the pith must have 

 been compressed (positive tension), the remaining tissues are stretched (nega- 

 tive tension). The epidermis is stretched most, the cortex less, and the wood 

 least of all, hence the epidermis is in a state of positive tension as contrasted 

 with the cortex, and the cortex as compared with the wood, or, in other words, 

 each layer is in a state of negative tension in comparison with that lying next 

 outside. The above table is only one example of such tensions ; in addition 

 it should be stated that the maximal shortening in the cortex observed by 

 SACHS was 5-9 per cent., and the maximal elongation of the pith 8-7 per cent. The 

 following experiments show that these tissue tensions are the result of different 

 degrees of turgidity in antagonistic tissues. 



If a stem be split longitudinally by two cuts at right angles to each other, 

 the four strips must necessarily curl up, so that the pith in consequence of its 

 efforts to expand will occupy the convex side. If the same experiment be made 

 with a hollow stem, such as the scape of Taraxacum, the same curvatures make 

 their appearance, because here also the peripheral tissues are in a state of 

 negative tension compared to those further in. If such a split stem be placed 

 in water, the curvature increases, because the cells are now saturated with 

 water, and are then able to follow out their efforts to stretch. Hollow stems 

 treated in this way often exhibit curvatures of such an extent that the strips 

 roll themselves up into spirals. If, on the other hand, the stems be placed 

 immediately after splitting in a plasmolysing solution, the original curvature 

 is undone, and the strips of tissue take on eventually a feeble curvature in the 

 opposite direction. If, on the contrary, the split stem has, first of all, lain in 

 water, and if it be then plasmolysed, it is not possible to remove the curvature 

 entirely, because it has now become fixed by growth. 



Let us now investigate the distribution of the tissue tension. Tissue 

 tensions make their appearance in the manner above described just as promi- 

 nently in vigorously-growing internodes as in petioles. As we pass from the 

 parts where elongation is taking place towards the growing point, the tensions 

 gradually disappear ; they are first observable where tissues are differentiated. 

 Further, they are almost completely absent from full-grown parts. The excep- 

 tions are the pulvini of many leaves, such as those of the Leguminosae and 

 Oxalidaceae. These pulvini are cushion-like swellings, which we are very 

 familiar with as occurring at the bases of the leaves of Phaseolus. A transverse 

 section through such a cushion shows a central vascular bundle surrounded by 

 parenchyma. If one cuts a longitudinal slice out of such a cushion, the effort 

 on the part of the cortex to expand in opposition to the tension of the vascular 

 bundle shows itself at once by the swelling of the originally straight upper and 

 under surfaces of the section. If the cortex be separated from the vascular 

 bundle, the former becomes concave inwards, and if it itself be halved longitu- 



