240 MR. C. DARWIN ON THE ACTION OF CARBONATE 
granules could be seen in the darker cells; and I suppose that 
this was owing to their being too minute to be visible. Occa- 
sionally, in the upper parts of the roots, the granules became con- 
fluent, and formed one or two small rounded masses of hyaline 
brown matter. The striped appearance sometimes extended from 
the tips of the finest rootlets close up to the stem of the plant. 
On a easual inspection it would be said that the longitudinal 
rows of brownish and of almost colourless exterior cells regularly 
alternated with one another; but on closer examination, two or 
three adjoining rows of cells were often seen to contain granules, 
and in other places two or three ordinary rows contained only 
colourless fluid. In one instance many adjoining longitudinal 
rows contained granules ; but the tendency to alternation was even 
here well shown, as the alternate rows differed in tint from in- 
cluding a greater or less number of granules. High up the roots 
the alternations often quite failed, as all the exterior cells con- 
tained granules. If a longitudinal row of cells with granules is 
traced up a rootlet, it is seen to be soon interrupted by one or 
more colourless cells ; but I have traced as many as 18 cells in à 
row all containing granules. So, again, a longitudinal row of 
colourless cells changes after a time into one with granular 
matter. Asa root thickens upwards, some of the longitudinal 
rows of cells divide into two rows ; and arow containing granules 
may divide into two such rows, or into one with and another with- 
out granules; and so it is with dividing rows of colourless cells. 
I could not perceive the least difference in shape or size, or in 
any other character, between the cells of the same rank which 
contained and those which were destitute of granules. 
Near the tip of the root it is the exterior cells which become 
charged, after immersion in the solution, with brown granular 
matter; and this often holds good with the cells of the root-cap. 
Higher up the root, the layer of cells formed by the alternating 
longitudinal rows with and without granules is sometimes bounded 
externally by a layer of empty cells, which, I suppose, had by some 
means been emptied of their contents, and were ready to be exfo- 
liated. Besides the exterior cells with and without granules, many 
separate cells in the parenchyma at different depths from the sur- 
face, and all or several of the elongated endoderm-cells surround- 
ing the central vascular bundle, are more or less filled with gra- 
nular matter, none of which cells contained any solid matter before 
the roots were immersed in the solution. 
