172 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



In the majority of instances, however, it was at least two days, generally 

 many more, before the edges were dead ; and in many cases they remained 

 fresh for prolonged periods, even for weeks. 



Nor is the result constant with the same kind of leaf. Some old 

 lilac leaves had but feeble power to nourish the parts in air when the 

 apical parts only were laid on water ; whereas leaves taken off the same 

 shoot with the apical part completely immersed, or else with the middle 

 part only in water, supplied the remaining parts sufficiently. 



As a contrast to the leaf of the Ipomoca mentioned above, another 

 leaf, placed with the upper surface of the apical half in water, nourished 

 perfectly the basal part in air, as well as a long stalk. 



Two leaves of borage were laid, one with the under surface, the other 

 with the upper surface of the apical parts in water, but they could only 

 nourish the mid-rib of the part in air ; the sides dried up as far as the rib. 



Both the upper and under surfaces of Digitalis purpurea nourished 

 the parts in air perfectly. 



In this and other corrugated leaves the water runs into all the minute 

 channels over the ribs and veins by capillary attraction, and thus 

 irrigates the entire surface. Garreau has noticed how these channels, as 

 well as the one very commonly occurring down the petiole, are particularly 

 advantageous for absorbing water. 



The conclusion I have arrived at is that the objection raised is really 

 of no consequence. In the majority of instances it is some days before 

 the margins dry up where the central part only is wet. Moreover, similar 

 leaves not kept wet always perish far sooner altogether. This shows that 

 even the leaves least capable of transmitting water laterally can do it to 

 some extent ; if the leaves be thick it is easily effected ; and with corru- 

 gated surfaces the transmission is not only within, but without as well, 

 so that the whole leaf becomes bathed with water, though the apex alone 

 may be actually in it. 



Now, when it is remembered that dew forms all over and on both 

 sides of leaves, they are never in this artificial condition of being wetted 

 only in part, at least at first ; but as the dew dries up in one part of a leaf 

 and transpiration has recommenced in sunlight, the above experiments 

 thoroughly establish the right to believe that an influx will be set up to 

 balance the renewed efflux caused by transpiration. 



G. On the Power of Absorption by detached Leaves laid on 

 the Surface of Water. 



Of the preceding experiments, the results were solely judged of by the 

 general appearance as presented to the eye. Such, however, clearly 

 proved that leaves can readily act as absorbing organs in the absence of 

 roots, not only so as to nourish themselves, but other leaves and buds on 

 the same shoot, especially if the stem be herbaceous. 



In the following experiments the leaves were left as stated below from 

 July 30 to August 3. They were all carefully weighed to the 5,000th 

 part of a gramme on the former date, then again at the latter. The 

 losses are reduced to percentages of the original weight of the specimens 

 respectively. 



