Doyle — Some Researches in Experimental Morphology. 423 



described in the rooted leaves. Here the outer layers became burst, and a 

 covering of periderm protected the exposed tissue. In all cases the cells of 

 the medulla were so packed with starch as to give the appearance when 

 treated with Iodine of a section of a potato tuber so treated. 



The leaves which were rooted and grown so for some time were of 

 interest. Only Pelargonium was investigated. In every case there was a 

 re-awakening of the old cambium, but the resulting activity was less than in 

 the B groups. Again, it was much greater near the petiole base than near 

 the lamina attachment. Its extent can be gauged from the portions of 

 bundles appearing in the next three figures. But most peculiar of all was 

 that without exception there was a very close demarcation indeed between 

 the portion above the ground and that below, because below the ground an 

 elaborate and considerable periderm was formed. 



Plate XXXIII, fig. 4, shows a complete view of the cross-section. The 

 small dark patches all over the slide are starch grains. But we see the 

 remarkable periderm surrounding tlie petiole. This periderm is shown on a 

 larger scale, Plate XXXIV, fig. 1. 



It is derived from the sub-epidermal coUenchyma layer of the petiole. 

 Thus, 1 is the epidermis intact ; 2 is the internal half of that coUenchyma 

 layer. Here it is very clearly standing out from the cortex. The" periderm 

 is definitely divided into two layers — («) an outer one with all tlie charac- 

 teristics of dead suberized cells, and giving the staining reaction of these ; 

 {b) an inner layer of living parenchyma with thin cellulose walls and round 

 in form like cortical cells. Close examination will show the presence in some 

 of these cells of minute starch grains. The cork cambium would seem to be 

 situated between the two layers — see the suggestion of cell-divisions at '6 — 

 cutting off true cork tissue on the outside and this parenchymatous tissue on 

 the other. The large starch grains in the medulla and the smaller grains in 

 the cortex are clearly visible. At 4 the small increase of secondary wood 

 is seen. 



But these leaf-cuttings showed other changes. More often on the above- 

 ground parts, but occasionally also usurping a portion of the petiole from 

 the periderm mentioned above, a fierce activity of medullary cells would 

 take place, resulting in the appearance of a split filled with cork. 



Plate XXXIV, fi.g. 2, shows this mad activity of these cells. A scrutiny 

 of the photograph shows the presence in them all of large numbers 

 of small starch grains. The whole gives the impression of a tendency towards 



8 y2 



