144 Professor Alexander Dickson on the 



ing. As seen in the transverse sections figured in Plate 

 XI., the cells of the innermost of these three layers — that 

 corresponding to the endodermis of ordinary roots — are 

 relatively small ; those of the middle layer are about 

 twice as large ; while those of the outer layer (which is 

 sometimes imperfect or almost absent) are of intermediate 

 size. Next comes tlie Pericamhium ; and then the Fibro- 

 vascular Zone. The fibro-vascular zone towards its peri- 

 phery consists of from 28 to 68 phloem tracts,* with the 

 spoke-like radiations of the xylem passing between them. 

 Internally, the xylem forms a continuous zone of mingled 

 prosenchyma and ducts. The xylem ducts are of largest 

 size towards the interior, and are smaller the nearer they 

 are to the periphery. The larger ducts exhibit scalariform 

 or dotted markings ; while the smaller ones, towards the 

 extremities of the xylem spokes, exhibit reticulated or 

 spiral markings. The centre of the root is occupied by 

 a rather large cylinder of thin-walled Medullary Paren- 

 chyma. As regards the general character of the root- 

 structure, it will be seen that, with exception of the 

 very remarkable development of the endodermis and the 

 somewhat collenchymatous character of the cortical 

 parenchyma, it corresponds with the ordinary type of 

 Monocotyledonous roots. 



In the course of my investigation, I was much struck 

 with the great dissimilarity in structural detail between 

 the roots from the plant in my own garden and those from 

 the plant in the Edinburgh Botanic Garden ; and this 

 induced me to apply to other establishments for further 

 root-specimens. The examination of these has led me to 

 recognise three types, — possibly of varietal importance, — 

 which are illustrated by the figures in Plate XI., drawn 

 to one scale, by help of the camera lucida. 



In Plate XI. fig. I, is represented the transverse section 

 of a portion of a root from the plant in my own garden. 

 The tissue elements are of small size. The endodermis 

 and xylem prosenchyma are relatively much indurated, 

 especially the endodermis, in which the cell cavities are 

 much reduced ; and even the pericamhium shares in the 



* These figures indi'-ate tlie smallest and largest immliers of phloem tract 

 that I have counted. 



