ACTION OF GRAVITATION ON GROWTH IN LENGTH. 851 



horizontal. In the position (a) the vertical arrangement of the tissues will be as 

 follows : — 



Cortex, 



Pith, 



Cortex : 



in this position the slice will always curve upwards. In the position (6) the different 

 tissues lie side by side in a horizontal plane, thus. 



Cortex, Pith, Cortex, 



and nearly the whole of both the upper and under surfaces will be occupied by the 

 section of the pith. In this position it often happens that no geotropic curvature is 

 exhibited, but accurate observation is rendered difficult by the unfavourable conditions. 



If a prism of pith be cut out of the growing region of a solid shoot, without any other 

 tissues being attached to it, and if this be placed for five or ten minutes in water so that 

 it becomes rigid and turgid, and be then laid horizontally in moist air or in water, one 

 end being fixed and the other free, no upward curvature takes place. 



2. Upcward Curvature of Grass-haulms. In the case of the stems considered in the 

 preceding paragraphs, the whole region of growth, which is of considerable length, is 

 geotropic ; hence the curvature is gradual and therefore also of considerable length, 

 and every portion of the stem which has completed its growth in the erect position has 

 become incapable of curvature. In the haulms of Grasses, on the other hand, the 

 capacity for curvature is concentrated at the nodes, the long internodes remaining 

 straight. Hence, a haulm possessing several nodes, if laid horizontally, will exhibit after 

 a short time a number of sudden angular curvatures at the nodes, between which lie the 

 straight internodes. If the oldest internode be fixed in a horizontal position, the third 

 or fourth internode will have assumed an erect position in from one to three days. It is 

 on this that the upgrowth of * layered ' Wheat depends. 



It is an especial peculiarity of the haulms of Grasses that they retain for a con- 

 siderable time the property of becoming erect, that is, that their under surfaces will 

 grow rapidly if they lie horizontally, after that they have ceased to grow while in the 

 normal vertical position. An abnormal position does not only aff^ect the growth of 

 the nodes of a haulm, but even causes it to recommence after it has already ceased ; 

 in this particular the motile organs of periodically motile leaves, e.g. those oi Phaseolus^ 

 resemble the Grass- haulm. 



These nodes which are thus capable of curvature are, as is well known, the basal 

 portions of the leaf-sheaths, which surround the base of the internode as a more or 

 less well-developed annular swelling of considerable thickness but of delicate succulent 

 structure. My observations were made upon Triticum, Dactylis, Glyceria spectabilisy 

 Andropogon niger and Zea Mais, in which plants the nodes are sufficiently large to 

 admit of tolerably accurate measurements. Portions of the haulms were cut off in 

 such a way that there was a node in the middle of each, connected with an internode 

 above and below ; the lower cut end was fixed laterally into moist sand in such 

 a way that the whole piece was horizontal : it was then put in a closed metal box, 

 the atmosphere of which was damp. The free internode was found to become 

 erect after two, three, or four days, according to the thickness of the node : sometimes 

 it became vertical, but more commonly oblique. It was easy to see that the lower 

 surface had el9ngated considerably. If the piece of haulm be now turned in such a 

 way that the convex surface of the node is uppermost, the concave side begins to 

 grow very vigorously; the node becomes straight, and appears uniformly elongated 

 on all sides. 



The convex lower surface of a sharply bent node appears smooth, transparent 

 and glistening, whereas the concave upper surface is dark, opaque, and rough in 

 consequence of transverse folds formed by its epidermis and parenchyma. A deep 

 indentation may often be observed in addition, so that it appears as if the node had 



3 I 2 



