I 



.8,,. THE FALLING OF LEAVES. 685 



early appearance of the separating layer. In the middle of June a 

 transverse layer of parenchyma dies and becomes suberised, completely 

 separating the parenchyma of the stem from that of the petiole, while 

 leaving the vascular bundles free. This evidently corresponds to 

 Mohl's round-celled layer, which takes on a brown coloration. Shortly 

 after, the layer of cells in the stem immediately below it becomes 

 meristematic, dividing centripetally by walls parallel to the layer. The 

 daughter-cells become suberised from above downwards, and a 

 layer of cork is formed which thickens more and more against the 

 lower border of the first suberous layer. This cork layer is in touch 

 with that of the stem. Thus a double suberised barrier separates 

 the parenchyma of the stem and leaf. A little later there appears in 

 the lower part of the petiole above the cork formation a layer of 

 secondary meristem, quite similar to and behaving like that just de- 

 scribed for the leaflet; it appears, however, much earlier, as it is so far, 

 that is, on the lower surface only of the petiole, perfectly developed 

 before the end of June. Leaves examined a month later were in the 

 same state, which is probably maintained without change till the 

 autumn, judging from Mohl's observations, which date from 

 October 4th. 



In the present case, leaf-fall was produced artificially by enclosing 

 the twig bearing the leaf in a damp box. For the first five days no 

 alteration was visible, but on the sixth the blue line produced in 

 staining a longitudinal section with iodine tincture, which marks the 

 layer of separating meristem, had progressed upwards, traversed the 

 vascular bundles, and reached half-way to the upper epidermis. On 

 the seventh day it had quite reached the latter as in the leaflet. The 

 central row of cells is absorbed, those of the upper and lower layers 

 becoming turgid, press on each other, causing the vessels and fibres 

 to snap, and the leaf falls. The parenchymatous cells of the end of 

 the bundles exposed on the scar become meristematic and form a cork 

 layer and also grow into the vessels, filling them with tissue (tylose), 

 so that the wound is soon covered by a continuous cushion of cork. 



The account of MM. Guignard and Van Tieghem suggests 

 several modifications of that given by Mohl : — 



1. The very early appearance of the separating layer in the 



lower region of the petiole. 



2. The similar early appearance of the suberous layer, the 



true nature of which Mohl did not appreciate, and also of 

 the underlying cork, which he merely states to have been 

 present at the commencement of his researches. 



3. The mode of separation which, according to Mohl, occurs 



without cell-destruction, the layers merely separating from 



each other ; whereas the two later observers speak of the 



absorption of an intervening layer. 



The French botanists also investigated the process in other leaves 



and leaflets. They found that all the leaflets fall as in Gymnocladus 



