244 MR. C. DARWIN ON THE ACTION OF CARBONATE 
dinal rows of cells eontained brownish matter; but this matter 
in several instances consisted of small dark-brown spheres, due 
apparently to the aggregation of granules. The endoderm-celis 
round the vascular bundle contained either similar spheres or 
granular matter. 
As many adjoining rows of cells on the surface of the roots of 
this plant had the same character, an excellent opportunity was 
afforded for observing the relation of the root-hairs to the celis ; 
and in several dissected roots it was manifest that, as a general 
rule, the hairs rose exclusively from the colourless empty cells; 
whereas none arose from those containing granules. Twice, how- 
ever, partial exceptions to this rule were observed: in one case 
the exterior walls of two adjoining cells, and in another case those 
of four adjoining cells, projected, so that they formed short blunt 
papille which included granules; and these papille exactly re- 
sembled nascent root-hairs. It is not, however, certain that they 
would ever have become fully developed. 
All the exterior cells close to the tip of the root in this case 
and in many others contained matter which was acted on by car- 
bonate of ammonia; and I was led by various appearances to ` 
suppose at one time that this matter remained in all the higher 
cells until it was consumed in some of them by the formation of 
the root-hairs. These consequently would arise exclusively from 
cells in which no granules would be deposited when they were 
acted on by the solution. In opposition to this supposition is 
the fact, first, that root-hairs could be seen beginning to be de- 
veloped from empty cells; and, secondly, that very many cells 
which were empty apparently had never produced root-hairs. 
Nor does this notion throw the least light on single cells in the 
parenchyma and on many cells, though not all, in the endoderm 
containing granular matter. 
With another Euphorbiaceous plant, Celebogyne ilicifolia, the 
immersion for 20 hours of its roots, or of thin sections of the roots, 
in a solution of 4 parts of carbonate of ammonia to 1000 parts of 
water produced a singular effect; for many separate cells in the 
parenchyma and those in the endoderm surrounding the vascular 
bundle assumed a pale or dark blue, and sometimes a greenish 
colour. As far as I could judge, both the granules within these 
cells and the cell-sap became thus coloured. Irrigation with sul- 
phurie ether did not affect the colour, though the many oil- 
globules in the cells were dissolved. 
