170 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



A similar specimen not wrapped up was perfectly dead in two days, 

 the previously juicy stem being now dry and collapsed. 



The long time during which the leaves remained green, of the first 

 described of these specimens of borage, clearly proves that the supply of 

 water must have been obtained through the epidermis of the internodes to 

 balance the transpiration. 



Symphoricarpus, or Snowberry. — A shoot had one internode wrapped 

 up, with four leaves beyond it exposed. After three days the lower pair 

 of leaves were dying, but the upper pair were fresh. After eight days all 

 the leaves were dead ; but the internode remained fresh several days 

 longer. 



A second and similar specimen had also four leaves exposed ; but they 

 were below the internode which was wrapped up. The order of decay was 

 in this case reversed ; the two lowermost or furthest from the wet inter- 

 node died first, those nearest to it last. 



The above are selected from a large number of experiments to 

 illustrate the fact that herbaceous internodes readily absorb moisture in 

 the endeavour to supply the leaves with water for transpiration, but that 

 the demand is usually much greater than one or a few internodes can 

 furnish : hence the leaves soon begin to die back from their apices to 

 their bases. In addition to such supply as they can for a time give to the 

 leaves, the experiments prove that moisture applied to internodes arrests 

 death and decay in the stems and axillary buds for variously prolonged 

 periods ; for efforts to develop axillary buds were frequently made, as well 

 as adventitious roots, these being apparently special instruments for 

 absorbing superficial moisture. 



3. On the Absorption by Leaves attached to Branches, and 

 their Power of Nourishing the Rest of the Leaves on 

 the Shoot. 



On July 23 a shoot of hazel, with a sub-herbaceous stem, had three 

 leaves lying with their lower surfaces only on water. The shoot .bore two 

 large and two small leaves sustained in the air. The whole shoot was 

 perfectly fresh and vigorous at the end of a week. On August 3 the 

 larger leaves began to die back from their apices, while the terminal 

 small ones were dead. Hence it was far from entirely perishing after 

 ten days. 



A similar specimen bad two large leaves with their upper surfaces 

 only lying on water, the remaining leaves as before in air. Like the 

 preceding, the whole kept perfectly fresh for the same time. The apical 

 leaves began to die about August 3, or after ten days. 



A similar specimen to these two, without water, was dead in two days, 

 the leaves being brown and brittle. 



Shoots of lime, elm, &c, treated as above gave similar results, 

 showing that the presence or absence of stomata is immaterial, the upper 

 surfaces of the above having none at all. 



These experiments entirely corroborate the results of Hales, Bonnet, 

 Baillon, Duchartre, Boussingault, &c, the general conclusion being that 

 the duration of life in the specimen thus treated depends upon the supply 



