Doyle — Some Researches in Experimental 3Iorphology. 417 



But, whatever our regrets might be, this digression — already much too 

 long — has shown us (of. Julius Caesar and Dagata) that, in spite of tlie 

 essential differences between the structure of the petiole described by Jannicke 

 and that one till recently under consideration by us, Jauuiciie was quite 

 accurate, liaving apparently used some such variety as the said Dagata. 



2. 'The Successive Microscopic Changes. 

 After so long a wandering we get on now to a consideration of the 

 microscopical changes that take place in the petiole after the grafting of the 

 sprout upon it. 



The first change is the re-awakening of the old dormant cambium in 

 many of the already existing bundles. This always appeared in that 

 aggregate of bundles which we mentioned as situate above and below the 

 central bundle, less frequently, however, in the lateral bundles, possibly due to 

 the fact that these bundles never made a satisfactory union with the vascular 

 cylinder of the grafted shoot, though this was not investigated. The 

 behaviour of the central bundle was interesting. Just as in tiie peripheral 

 ones the dormant cambium became very active, but as well there appeared on 

 the xylem-free side of that bundle a series of cambial divisions (Plate 

 XXXI, fig. 8). 



Here we see a-a the band of cambial cells appearing on that side of the 

 phloem ring abutting on the cortex (/. h-b is the internal parenchymatous 

 mass already often referred to. c-c is that portion of the phloem which lies 

 against the primary wood. 



This layer very soon multiplies to form a typical cambium band (a), 

 Plate XXXI, fig. 4. (b) is the cortex, (c) internal parenchyma. 



Though the figure is somewhat hazy, it shows clearly that this cambium 

 layer has been derived not from the surrounding cortex, but from the phloem 

 band. 



Both this cambium and the bands of cambium which became active again 

 elsewhere build up secondary wood. 



Plate XXXII, fig. 1, shows this extremely well. 



(a) is the primary wood, {b) is the wood formed from the re-awakening of 

 the original cambium. Its cliaracteristic close texture is very visible, 

 consisting of large vessels scattered in a compact mass of thick-walled 

 tracheides. To what extent fibres are developed has not yet been investigated. 

 (c) shows the beginning of secondary wood of a similar nature on the 

 formerly xylem-free side. 



The portions (a) and (b), with the cambium and phloem appertaining 

 thereto, are typical of the larger peripheral bundles at this stage. 



