in the septal glands of Narcissus. 63 
gland-cavity in the first instance takes place through the 
cell-walls (which are not cuticularised) without any rupture, 
splitting away of the cells of epithelium from one another, or 
mucilaginous degeneration, and must therefore be supposed to 
result, in the first instance at least, from the direct activity 
of the protoplasm in the secreting cells. 
Finally, I should wish to call attention to the close analogy 
between the results deduced from these observations and those 
of W. Gardiner 1 on the secretion of mucilage in the hairs of 
Blechnum occidentale and Osmunda regalis , and especially to 
his remarks on the similarity of the process of secretion in 
its general features by animal and vegetable protoplasm. 
I think also that the formation of sugars in this manner may 
be regarded as comparable to the formation of cellulose from 
microsomata 2 , and of starch from amyloplasts 3 in its general 
nature, but in such a comparison it must be remembered that 
the formation of solid products, such as mucilage, starch, 
cellulose, etc., from specialised portions of the protoplasm 
admits of direct proof, whereas it is hardly possible to obtain 
more than indirect evidence in the case of soluble bodies, such 
as sugars. 
1 Annals of Botany, vol. i, No. i. 
2 See Vines, Phys. of Plants, p. 25-26, Cambridge, 1886. 
3 See Vines, loc. cit. p. 26 and 180. 
