92 



JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



part of the stem in the air, the leaves helow become dissected by the 

 total arrest of the intermediate tissue, while those in air are com- 

 pletely formed. The suggestion was that the living protoplasm of the 

 stem was injuriously saturated, so that it was too weak to make a 

 perfect leaf. Acting on this "working hypothesis," he made the 

 water denser by dissolving certain nutritive salts in it. This set up 

 ** osmosis," the result being that his assumption proved to be correct. 

 Water was withdrawn from the stem, and the subsequent leaves were 

 perfectly formed, though under water.* 



Nature's experiments may frequently be seen. Thus if Ranunculus 

 ^ trichophyllus , which has no floating leaves, grows very thickly, several 

 branches may be forced up into the air. When this is the case, the 

 whole of the tissue changes and becomes adapted to air from the water- 

 level upwards. If the pond dries up, the Crowfoot grows just as well 

 when rooted in the mud. 



If it be R. heterophyllus , then both the submerged and floating types 

 of leaf are found in air. If the seeds be sown in a garden, they all 

 come up retaining these two forms of leaf, acquired during many 

 generations of an aquatic life. 



The Sheathing Base of the Leaves of Aquatic Plants. — Another 

 feature common to aquatic and semi-aquatic plants, which often grow 

 tufted because the stem is arrested, is to have a broad base to the leaf- 

 stalk, thus more or less sheathing the stem. This is due to the fact 

 that while an ordinary leaf-stalk or petiole of a plant living in dry 

 conditions usually receives one or three strands out of the woody 

 cylinder of the stem, a good many more are supplied to the leaves 

 of aquatic plants; because the bundles or "strands" have 

 become more or less separated in the stem ; so that the petiole, so to 

 say, widens out to receive them. This occurs in many of the 

 Ranunculaceae and Umhelliferae, a portion of which families are now 

 of an aquatic habit, and probably several others were so formerly. 



In some plants a rather different procedure takes place. There is 

 a tendency to form a horizontal nodal plexus of interlacing fibro-vascu- 

 lar bundles connecting the vertical strands ; from this several pass off 

 into the leaf -stalk, which widens into a sheath. This may be seen in the 

 Lesser Celandine {Ranunculus Ficaria) and Winter Aconite {Eranthis 

 hyemalis), Marsh Marigold {Caltha palustris), Docks (Rumex), Fennel 

 (Foeniculum), and other umbelliferous plants. 



The Sequence of the Forms of Some Aquatic Leaves. — In the 

 forms assumed by leaves of aquatic plants there is a kind of sequence 

 in their development. Thus of the Water-lily family, in Victoria 

 regia, being an annual, it is especially observable. Tracing the 

 development from germination, the first leaves are greatly arrested, 

 consisting of sheathing petioles only, comparable in form but not 



* Bot. Gaz. vol. xxxiv. p. 93. This experiment agrees with the results of 

 " Water Culture." Many plants ai-e grown in water for experimental purposes, 

 the water having been first supplied with nutritive substances. The tap-root 

 is no longer arrested, but grows with secondary rootlets, &c. 



