Flood — Exudation of Water by Colocasia antiquorum. 509 



canals. This was done by sealing the petiole of a leaf into a bottle with 

 plasticene, and exhausting the air from the bottle. The tip of the leaf was 

 kept dipping into a mixture of gelatine and Indian ink, which was solid at 

 the temperature of the laboratory, and liquid at a temperature somewhat 

 higher. It was kept liquid during the experiment by being surrounded with 

 hot water. The leaf was kept warm ; the petiole dipping into hot water, in 

 order to keep the conditions near to those obtaining in the hothouse from 

 which it was brought. The result of this experiment was that no gelatine 

 mixture was drawn in through the pore. Then the tip was cut off and the 

 experiment repeated. This time the mixture mounted easily up the canals. 

 This seemed to indicate the presence of a membrane ; but the objection to this 

 experiment was that the passages leading from the pore might be so small that 

 viscosity or capillarity might prevent the mixture from getting through them. 

 To reduce the viscosity, a mixture of water and Indian ink, wilhout the 

 gelatine, was used, under the same conditions, but, like the gelatine, it did 

 not get through from the pore into the canals. 



Then, owing to the fact that when the plant is in action the water does 

 not rest in the depression, but is either shot off or runs down to the extreme 

 tip of the leaf to form a large drop, the presence of wax or of some greasy 

 substance was suspected. This would prevent the water and Indian ink 

 from wetting the surface of the depression and entering- the pore. To dissolve 

 off the wax, the tip of a fresh leaf was plunged for a few minutes into ether, 

 the ether was washed off with spirit, and the spirit in turn removed by 

 washing with water. This was done as quickly as possible, in order to 

 minimize the injury due to penetration by the ether and spirit. The leaf was 

 sealed into the bottle as before, and the experiment carried out under the 

 same conditions. It was then found that a mixture of Indian ink and water 

 went through the pore and up the canals. The inference drawn from this 

 experiment was that there was no continuous membrane under the depression, 

 as a colloid pigment could pass through the pore and up the canals. 



In order to verify this result further, a piece of the blade of a leaf with 

 the tip attached was smeared with glue and rolled round a small cork in such 

 a way that the cut edge of the blade projected at one end of the cork, and the 

 tip hung free at the other. A hole was bored in a large cork, and the small 

 cork with the leaf glued into it. The large cork was fitted into the neck of 

 a receiver, so that the cut surface of the leaf -blade projected into the receiver, 

 and the apex of the leaf hung free. When the corks were satisfactorily fitted 

 so that the only possible outlet for water was down the canals and out by the 

 pore, the receiver was partly filled with water and a fresh surface of the leaf 

 cut under water. The receiver was then filled up till the water exerted a 



